CN101819609A - Be used to design, follow the tracks of, measure, predict and optimize the system and method for data communication network - Google Patents

Abstract

A kind of system and method that is used to design, follow the tracks of, measure, predict and optimize data communication network.The method of this analysis and adjustment cordless communication network may further comprise the steps: the computerized model that produces or use the cordless communication network in the physical space with computing machine or server; At described computing machine or server, receive measurement data from one or more measurement collector or agencies that are arranged in described physical space; One or more performance index at described computing machine or the described cordless communication network of server prediction; And the setting or the configuration that change at least one assembly of described cordless communication network based on the instruction of sending from described computing machine or server.

Description

Be used to design, follow the tracks of, measure, predict and optimize the system and method for data communication network

The application be enter the China national stage on March 25th, 2003, the applying date is September 21 calendar year 2001, application number is 01816269.X (PCT/US01/29419), denomination of invention is divided an application for the patented claim of " being used to design, follow the tracks of, measure, predict and optimize the system and method for data communication network ".

Technical field

The present invention relates to communication network field, the particularly design of data communication network, measurement, visual, prediction and performance optimization.The model of the element by using specific physical environment in accurate scene and composition data network, to the data communication network predict, method and system visual and performance optimization can be used for these data networks are related to, measure, monitor, arrange barrier and improve.

Background technology

Communication network is used for information is sent to the another location from a position.This information often is the form of speech, video or data.For transmission information, communication network is decomposed into a series of numerals with message.These numerals have been described some preordering methods of use and how to have been come structure information.For example, the digital sample that these numerals can representation signal voltage is used for loudspeaker with this voltage, makes its sound that can reproduce speech, as shown in Figure 1.In this example, information is the speech information that transmits by communication network.

The process of expression information can be an analog or digital.In the analogue communication network, the message of transmission is the numeral that changes continuously.In digital network, message by discrete, with certain hour at interval rather than the continually varying numeral represent.Each represents signal by an independent numeral at interval.This numeral can be converted to binary mode, so that whole message can be expressed as 0 and 1 finite population.Each binary digit in the message is called as a bit.These bits are transmitted and are translated into message by receiver.The scale-of-two of signal and digital model are as shown in Figure 2.

Digital communications network is a kind of certain kinds shape in the communication network, and it by wired or wireless network, to receiver transmits the numerical information that with bit or byte (groups of 8 bits) form shows from sender unit at indoor or outdoors.Although conceptive very simple, be complicated from an A to the device that a B sends data, and change in force.Hundreds of agreement, hardware, device, software engineering and program are arranged with solving the problem that how correctly transmits data effectively.Because the time variation of the channel of this complicacy and data communication network and operation thereof is to Effect on Performance, the definite performance of a specific data communication network is very difficult to prediction even measures.

Data communication network can be divided into circuit switching or packet switching network.This network of two types all uses channel transmission information.Channel is the communication port between the user in the communication network.A channel can comprise many independent hardware units, and the specific route between sender and the receiver.In circuit-switched network, the channel by unique reservation comes transmission information.Network channel is reserved, be used for unique transmission, once send whole bits.An example of above-mentioned situation is to use facsimile recorder to send file.In this example, facsimile recorder is a pixel with the image transitions of file.Each pixel is that on the paper is little, the rectangular area of point-like.Each pixel is black or white.Data to be transmitted are that the expression every bit is a series of bits of black or white.When being ready to message (in this example for document image) when a facsimile recorder is sent to another,, set up the transmission that a telephone line is specifically designed to these data by on common older telephone systems (POTS) communication network, carrying out call.Telephone wire is monopolized by facsimile transmission, makes it become Circuit-switched transmission.After setting up connection, all data send to second with the bit stream of an independent length from first facsimile recorder.In this example, transmit described bit with the tone of different frequency on the telephone wire.High-pitched tone is represented " 1 ", and low pitch is represented " 0 ".The facsimile recorder that receives is translated into data bit by the sequence with high low pitch and receives bit in the message.Then, by drawing the point of black in the position of data bit indication, the duplicate that the facsimile recorder of reception can the reconstruct source document.

Packet switching network is another type of data communication network, wherein transmits all data bits with a lot of little data-bit-group that are called as grouping, and it is sent to the another location from a position individually.Grouping is self to have comprised the part that also has the full message of footnote by header, data bit, sometimes.Grouping comprises the information of header and footnote, and this information allows data transmission network correctly to send grouping, and makes it know that the data in the grouping are parts of which message.Usually by the identifier mark, network uses this identifier to forward the packet to correct receiver to header.The information of header and footnote is usually used in grouping and other groupings are rearranged, and with the reconstruct origination message, and checks in the transmission packets whether error code takes place.By removing header and footnote, the data bit in the grouping is added title and collects again is origination message, and all grouping compilations that receiver will receive are origination message.

According to the transmission packets mode, packet switching network is divided into according to connection and connectionless.In according to the network that connects, during transmitting, each uses predefined network channel.Although can comprise a plurality of groupings in the transmission, the route between having set up from the sender to the receiver, thus all can directly be sent to receiver immediately in groupings that this channel sends.In connectionless network, in many transmission, side by side transmit grouping by shared channel.In this case, each grouping needs the identifier that the receiver address is provided.This address is understood by communication network, will divide into groups suitably to be sent to correct receiver to allow it.Because each grouping can transmit respectively, and with interweave from being grouped on the time of other transmission, so when using network resource shared, it is more effective not have the transmission of connection usually.

Connectionless, be to transmit according to an example of transmitted in packets at the file that is connected to according between two computing machines of the Ethernet of Internet protocol (IP).In this example, the file that will send at sender is decomposed into suitable grouping, and with IP address mark, this address is the identifier that network is used for forwarding the packet to correct receiver.The computing machine of reception then will divide into groups to be sent to from the computing machine that sends.Ethernet can by the stream of packets of control from each target, be supported to transmit from a plurality of files of the many various computing machines that use consolidated network in the mode of sharing.Then receiver will divide into groups the compilation be the correct duplicate of source document, finish this transmission.

All data networks all use some forms of communication protocol, come the transmission and the reception of Rule Information.Agreement is a rule set, and all hardware in the communication network and software all must be observed this rule to allow to carry out suitable data communication.In the message exchange in the world, there is hundreds of kind agreement to use now.Such as some agreements of transmission control protocol (TCP) or User Datagram Protoco (UDP) (UDP), defined the mode of accesses network in these agreements.Such as other agreements of Internet protocol (IP) or file transfer protocol (FTP) (FTP), message and grouping how have been defined to formaing, send and receiving.

Whether normally all data communication networks can come its efficient of assay and performance with certain form, and confirm network operation.In order to estimate the function of these data networks, use certain performance standard.These performance standards include but not limited to: throughput, bandwidth, service quality, bit error rate, packet-error-rate, the frame bit error rate, packet loss, packet awaits time, return interval, propagation delay, transmission delay, processing delay, queueing delay, network capacity, packet jitter, bandwidth-delay product and switching delay time.Each performance standard has been stipulated different performance parameter in the data communication network.Will be discussed in more detail below these standards.

Link is the part of the transmission path of the message between sender and receiver in the data communication network.Network connects and comprises independent device usually, is used for network packet is relayed to receiver from sender.This means that network connects some actual transmissions that can comprise between original sender and intended receiver.Each independently relaying be called a link.Typically a whole network connects and comprises some links.Can measure the performance index of each link.

Throughput is the measurement to data volume, and these data can transmit between two positions in data network, do not comprise header, footnote or routing iinformation bit.Usually be unit with bits per second (bps), can limit the hardware, software, fire wall and the above-mentioned every any combination that have constituted the connection between the sender and receiver in the data communication network.The definition of the bandwidth of data communication network is similar to throughput.Bandwidth is the speed of the raw data of particular communication network support, and is slightly higher than throughput usually.For example, an ethernet link can be the 10Mbps bandwidth, and the measurement result of actual file transmission shows, use the actual speed rate of data of two intercomputers transmission of same link to have only the throughput of 6.8Mbps, referring to Peterson, L.L.and Davie, B.S., Computer Networks:A Systems Approach.San Francisco:MorganKaufmann Publishers, 2000.

Service quality (QoS) is to be used to describe with the term of a certain amount of allocated bandwidth to the network of particular network sender.Described network allows the certain bandwidth of transmission requests.Then, network will determine whether can guarantee this bandwidth.As a result, network program has a reliable bandwidth that can adapt to easily.When the service quality that measure to connect, answer the actual bandwidth of bandwidth that the comparing cell requirement provides and the bandwidth of different requests.

Difference between the bit that Fig. 3 represents, grouping and the frame.Bit, grouping and frame for data communication network have defined the different bit error rates.Bit is the core of grouping and frame.Bit is the real messages in transmitted over communications networks.Grouping comprises data bit, packet header and grouping footnote.Packet header and grouping footnote are added by communication network protocol, are used for guaranteeing that data bit sends to the tram of data network, and are correctly deciphered by receiver.Packet header and grouping footnote also are used for guaranteeing grouping transmission correctly, and detect error code when error code occurring.Frame is the bit sequence with a definite form and form, and described form makes receiver know the beginning and the end of a frame.Bit error rate is bit shared number percent in the bit number of transmission of being receiver or no show receiver improperly.Packet-error-rate or packet loss are to arrive shared number percent in the total number packets that is grouped in transmission of receiver or no show receiver improperly.The frame bit error rate is the frame shared number percent in the frame sum of transmission that arrives receiver or no show receiver improperly.

Use some terms to limit the quality of the time delay of particular network incident, can represent to represent second with chronomere.The packet awaits time is that grouping is sent to the needed time of receiver from sender, and return interval, (RTT) was transferred to receiver with grouping from sender, and certain affirmation is turned back to the required time of original sender from receiver.Propagation delay, transmission delay, processing delay and queueing delay are described the time that needs in the various process of grouping transmission.Add up by propagation delay, transmission delay, processing delay and queueing delay, obtain packet awaits time and return interval that network connects one way in the network or backhaul.Propagation delay is that the required time of physical distance between sender and receiver is crossed in grouping.Transmission delay is last the required time of bit arrival that reaches same grouping from first bit of grouping.Processing delay is meant that the sender end is further divided into independent grouping required time with data-message, and receiver end required time of reconstruct complete data message from packet.Queueing delay is meant waits for that other transport process discharge the required stand-by period of shared resource.All these time delays for the different aspect of evaluating data COMMUNICATION NETWORK PERFORMANCES all of great use.

Two other network performance standard is packet jitter and bandwidth-delay product.Packet jitter be time of arriving of grouping with the variation between the time of regular speed arrival, typically measure second with chronomere.Bandwidth-delay product is before the actual arrival of the bit of first transmission receiver, the bit number that sender can send.Multiply each other by bandwidth and to obtain bandwidth-delay product packet awaits time of link and same link.

In radio data network, when the user shifts out an access point scope and enters in another access point scope, switch.In this situation, first access point must be delivered to the responsibility that transfers data to the wireless user second access point.Be for making the wireless user be connected to another access point from an access point switching time, and an access point and another access point are coordinated required time quantum.

Now developed and be used in the life-span of data communication network the software program of the performance statistic of measurement data communication network and hardware unit.Below the concise and to the point some of them related tool relatively more commonly used of describing.

A large amount of command-line tools can be used for making the computer user to measure approximate network performance in the connection fast.Many order line programs are widely used in Windows, UNIX and Macintosh operating system, on the diagnosis of data network and row's barrier a little effect are arranged also.The example of these order line programs comprises ping and Traceroute.Utility command line program ping can measure the approximate data stand-by period between the different pieces of information network equipment, and confirms whether available the network between stream oriented device connects.Network connects and comprises the autonomous device that network packet is relayed to receiver from sender usually.This means that the network connection can comprise some actual transmissions between original sender and intended receiver.Each independently relaying be called as link.Typically a whole network connects and comprises some links.Therefore, use Traceroute, can be between sender and receiver determine a possible path in from the relay to the relay, thereby awareness network transmits employed definite link.In addition, use Traceroute, can measure and cross the required time of each link, can also determine the link that may normally not move.

It is more accurate that a lot of command-line tools that are not contained in the operating system also are used to finish, but still be the network measure task that is similar to.The example of these instruments comprises: ttcp, Tcpdump.Ttcp represents test tcp Http:// www.pcausa.com/Utilities/pcattcp.htm, it is the free program of writing for the BSD (SuSE) Linux OS originally, but can be used for UNIX and Microsoft Windows operating system now.Ttcp is a basic point-to-point throughput process of measurement, and it makes size, different low level TCP or the UDP option that the user can controller buffer, and the definite data of control transmission.

Tcpdump is by in the one group of instrument that is called as the detector that divides into groups and a simple program that comes.The grouping detector allows the network manager to check the content that comprises header and footnote information on the network in the actual packet.The grouping that Tcpdump allows customer inspection main frame (although need not to be for this main frame predetermined) to receive, and show that all meet the header of the configurable form of user.Owing to allow the user directly to check definite network traffic, so Tcpdump is the useful tool that network connects row's barrier.

Pathchar is a UNIX order line program, it is by the size of the test packet that change to transmit, can the measurement data communication network between each network trunk device (as: router, hub, switch) throughput and measure stand-by period of the grouping transmission of heterogeneous networks point.The function and the Traceroute of this instrument are closely similar, and it has increased the function of measurement throughput (though right and wrong are direct), and not only measures the stand-by period.Pathchar only is subjected to the restriction of the network hardware in the link that its measures.This program needs hub, switch or computing machine to transmit an affirmation to test packet.This means can not be by the pathchar independent measurement such as the hidden link that does not transmit affirmation of Ethernet bridge.

Some company's production network measures, monitoring, tracking and predictor.Introduce the program that some of them are generally used below.Selected instrument is used for illustrating network performance measurement and equipment tracking in the prior art.

The netViz that netViz company produces is a viewdata library, and it allows the physical location and the logic placement of network administrator track network equipment.This program allows setting, position and the structure of user's fan-in network, and follows the tracks of the equipment in your network.This instrument can still can not be followed the tracks of described equipment with three dimensional constitution with data storage in the two-dimentional geographical map or floor planning figure of a building.This instrument can not provide network test, measurement or monitoring function, does not also support the communication prediction or the performance visualization of data communication network.It is a database that is used for the equipment accurate tracking.

(former Ganymede Software Inc.) has produced network monitoring and the forecasting tool of a kind of Chariot of being called in NetIQ company.Chariot can measure throughput and a lot of other network statistical datas of all network type, operating system and agreements commonly used now.This program uses a server and some little Agents (agents) to collect data.Server is checked at interval by network installation each Agent on subscriber computer with certain hour, when depositing the result in server, is used them to detect network characteristic.These Agents can be measured the network that is connected on server or another network and connect, and can simulate the use pattern that the business model of any network program and any one or a plurality of supposition user are wished.This program can also be used measurement data prediction network traffic and condition.

Visonael company (former NetSuite Development company) has produced some network traces and measurement products, comprises NetSuite Audit, Design and Advisor.These software products can detect the network equipment in the use automatically.Then, the information that these information of same are manually imported can be placed together the physics or the logical diagram of network.Visonael also provides a kind of product, and it determines network suitably configuration, and can be change and the upgrading on your the network recommended structure.This software product can not be predicted or measurement performance under the mode of ad-hoc location, can not predict the performance according to wireless data communication network.

SAFCO technology company (being the part of Agilent company now) has invented some wireless data measurements and prediction product recently.A kind of product of SAFCO is called DataPrint, and it is used for measuring the various data performance parameters of mobile phone data network.Their product

Also support wireless data transmission in the wireless telephone network in the effect Analysis aspect total volume and the service quality.

Wirless Valley Communications, Inc. have invented a new notion, are called SitePlanner, and it can measure the also environment particular network performance of tracking data communication network in the physically accurate three-dimensional model of environment.SitePlanner uses the software module of LANFielder by name to measure any throughput, packet awaits time and packet-error-rate that is connected in any wired or wireless network in Internet protocol (IP) data communication network.In addition, SitePlanner allows a whole network is modeled as a precise analytic model physically, thereby can measure accurately and performance prediction in the mode of ad-hoc location.SitePlanner also allows the logic distribution of network is stored simultaneously with physical distribution.This instrument also materials used list form logic of storing arbitrary communication network connects and the ad-hoc location model.

Except network measure and equipment management tool, a lot of researchs have also been carried out in wireless data communication network performance field.The work that belongs to field of the present invention has been represented in the research of introducing below.

Xylomenos has studied at some fixing IEEE802.11 wireless lan networks with Polyzos and has been connected the UDP of transmission and the performance of TCP grouping, see Xylomenos, G., Polyzos, G.C. " TCP and UDP Performance over a Wireless LAN " Proceedings ofIEE INFOCOM, the throughput restriction that is caused by software executive problem and operating system shortcoming is paid close attention in 1999. these researchs.The researcher uses order line program ttcp, the Tcpdump of themselves modification and the test of throughput that nstat finishes UDP and TCP under Linux.Between three fixed positions, carry out all and measure, and pay close attention to the type (PCMCIA or ISA) of WLAN card and final user's computer hardware (as have 48MB RAM Pentium150, with the Pentium200MMX with 64MB RAM).Researcher's conclusion is to recommend to change the execution of procotol and improve linux operating system.Measurement result is not considered the influence of the variation of different physical locations or radio communication channel to the network throughput.

Maeda, Takaya and Kuwabara have published the measurement of WLAN performance and the validity (Maeda of ray trace technological prediction WLAN performance, Y., Takaya, K., and Kuwabara, N., " Experimental Investigation of PropagationCharacteristics of 2.4GHz ISM-Band Wireless LAN in Various IndoorEnvironments; " IEICE Transactions in Communications.Vol.E82-B, No.10 Oct.1999).Tracking measurement in a little firing frequency (RF) of being controlled by environment shows that the throughput of WLAN distributes relevant with BER with the delay of wireless channel.The researcher do not introduce any by by ray trace technological prediction to delay distribution profile output in the method for accurately predicting bit error rate or throughput.

Duchamp and Reynolds introduced the IEEE802.11 WLAN, for the measurement result of the grouping throughput of different distance, see Duchamp, D., and Reynolds, N.F., " Measured Performance of a Wireless LAN, " Local Computer Networks, 1992.Proceedings., 17 ^ThConference on, 1992.These measurements are to finish in an independent passageway.Therefore, these measurement results can not be measured representational environment.Research does not have to propose to be used for predicting their result's model, does not attempt to confirm the computer forecast technology of any kind of yet.

Bing has also proposed the measurement result of the performance of IEEE802.11 WLAN, see " Measured Performance of the IEEE 802.11 Wireless LAN; " LocalComputer Networks, 1999.LCN ' 99.Conference on, 1999.Bing has introduced for the delay of various wireless lan infrastructures and the measurement result of throughput, and according to showing the throughput and the time delay of theory.But the result provides as optimal results.All measurements are at wireless channel may influencing under the minimum mode of total throughput and time delay to be carried out.Therefore, the result is the upper bound of best possible outcome, and can not expand the forecasting techniques of the WLAN performance that is used for ad-hoc location.

Hope and Linge use measurement result, by the required parameter of wireless lan network area coverage in the Okumura Model Calculation prediction outdoor environment.The researcher uses standard IEEE 802.11 WLAN modulator-demodular units to carry out outdoor measurement, calculate the required parameter of Okumura model, and provided the result, see Hope, M.And Linge, N., " Determining the PropagationRange of IEEE 802.11 Radio LAN ' s for Outdoor Applications; " LocalComputer Networks, 1999.LCN ' 99.Conference on, 1999.Use these results, can calculate outdoor area coverage.But the result does not allow throughput and the performance of stand-by period in the user in predicting WLAN.

List following relevant patent below:

Patent?No.5,491,644?entitled“Cell?Engineering?Tool?and?Methods”filed?by?L.W.Pickering?et?al.；

Patent?No.5,561,841?entitled“Method?and?Apparatus?for?Planning?aCellular?Radio?Network?by?Creating?a?Model?on?a?Digital?Map?AddingProperties?and?Optimizing?Parameters，Based?on?Statistical?SimulationResults”filed?by?O.Markus；

Patent?No.5,794,128?entitled“Apparatus?and?Processes?for?RealisticSimulation?of?Wireless?Information?Transport?Systems”filed?by?K.H.Brockel?et?al；

Patent?No.5,949,988?entitled“Prediction?System?for?RF?PowerDistribution”filed?by?F.Feisullin?et?al.；

Patent?No.5,987,328?entitled“Method?and?Device?for?Placement?ofTransmitters?in?Wireless?Networks”filed?by?A.Ephremides?and?D.Stamatelos；

Patent?No.5,598,532?entitled“Method?and?Apparatus?for?OptimizingComputer?Networks”filed?by?M.Liron?et?al.；

Patent?No.5,953,669entitled“Method?and?Apparatus?for?PredictionSignal?Characteristics?in?a?Wireless?Communication?System”filed?by?G.Stratis?et?al.；

Patent?No.6,061,722?entitled“assessing?Network?Performancewithout?Interference?with?Normal?Network?Operations”filed?by?W.J.Lipaet?al.；

Patent?No.5,831,610?entitled“designing?Networks”filed?by?D.L.Tonelli?et?al.；

Patent?No.5,821,937?entitled“Computer?Method?for?Updating?aNetwork?Design”filed?by?Tonelli?et?al.；

Patent?No.5,878,328?entitled“Method?and?Apparatus?for?WirelessCommunication?System?Organization”filed?by?K.K.Chawla?et?al.

In patent 09/352,678,09/221,985,09/318,841,09/318,840 and other inventions of quoting previously in an existing product describing, SitePlanner, because this product can be predicted the influence of wireless frequency directly related with any communication network in any physical location, so it is useful for design, measurement and the optimization of communication network.That is, use the physical distribution of any communication network and the information of hardware configuration thereof, prior art can provide signal intensity density (RSSI) that expectation receives, signal to noise ratio (snr), receive power level, best server and power location of equal value and the visualization display of the useful parameter of other voice and data networks relatively any modeled physical location.These statisticss can be predicted the forward link (from the sender to the receiver) or the reverse link (replying from the primary reception machine to original sender) of wireless network.The ad-hoc location characteristic of these predictions is directly interpreted as the quick and useful visual of communication network quality.Yet, prior art is not considered suitably to complicacy modeling (as prediction), these complicacy are probed into operating performance parameter value how to determine real network, and these parameters are subjected to the influence of multichannel propagation, the many interfering datas transmission from multiple source, signaling protocol, equalization methods etc. simultaneously.Because different modulating detuner manufacturer has different and often is the method that proprietary being used to alleviates and handle multipath, multiple access interference, protocol type, grouping size and noise, so the standard of bit error rate, data throughput, delay and service quality of predicting in the three-dimensional physical model in the environment of the ad-hoc location of reality is the unusual task of difficulty, and still unresolved so far.That is prior art has shown how to measure, show and predict basic communication standard, but can not be provided for predicting the prediction algorithm that the significant data of large-scale network performance parameter is determined in mode reliable, ad-hoc location.In brief, owing to must consider possible physics, the electric and logic factor of all component in the network, and produce the factor in the wired or wireless channel of real network performance and the complicacy brought, therefore do not draw such wireless network performance prediction engine in the prior art, it can consider accurately modeled three-dimensional physical environment and the knowledge of using certain components to distribute, and described engine neither be conspicuous.

Not any invention of accurately prediction three-dimensional of explanation, ad-hoc location network performance standard in the paper of in communication network field, having delivered.Above-mentioned paper, Maeda, Y., Takaya, K., and Kuwabara, N., " Experimental Investigation of PropagationCharacteristics of 2.4GHz ISM-Band Wireless LAN in Various IndoorEnvironments; " IEICE Transactions in Communications.Vol.E82-B, No.10 Oct.1999 has illustrated prediction wireless channel delay transit, the content relevant well with throughput with prediction, but in fact described method can not predict throughput or any other network performance standard.Some DESCRIPTION OF THE PRIOR ART with the performance of certain physical precision with two-dimensional approach tracking network equipment, but these products do not have the similar or different physical environments of prediction (as install) future network performance ability.Many products can Measurement Network performance standard, but be not used in the three dimensional representation that the physics according to assembly of prediction network performance parameter is installed in the prior art.And, do not exist yet can directly measure in mode three-dimensional, ad-hoc location, the network performance standard of the equipment of tracking data communication network, predicted data communication network and make visual instrument of network performance standard or invention.

And, there is not to use the existing measurement result of existing network in the prior art yet, or the distribution details of the ad-hoc location of specific components in the use data communication network, come that network parameter to complexity carries out accurately, the invention of performance prediction ad-hoc location, three-dimensional.In addition, do not use agent system in the prior art yet, and with results that specific three-dimensional position and those are measured Agents be applied to create in identical, the similar or different physical environments future network performance the three-dimensional prediction model, come from the network system of reality or subsystem, to measure independently the network performance parameter of ad-hoc location.In addition, prior art does not have to set up the hierarchy system that is used for measuring and predicting in three-dimensional ad-hoc location mode network performance measurement and prediction engine yet, this system has the ability of Measurement Network performance parameter, and produces the ability of the engine of the measurable network performance parameter that can be shared by the remotely predicting engine.

The present invention with a kind of be not that conspicuous mode has been expanded prior art, provide the wireless specific performance properties standard of communicating by letter to be also referred to as the visual of performance parameter and measurement with ad-hoc location, three-dimensional accurate way, as throughput, bandwidth, service quality, bit error rate, packet-error-rate, the frame bit error rate, packet loss, packet awaits time, return interval, propagation delay, transmission delay, processing delay, queueing delay, network capacity, packet jitter, bandwidth-delay product and switching delay time with cable network performance prediction, significant data.The invention provides the distributed measurement technology of the novelty that is used to measure above-mentioned performance parameter.In addition, created the Forecasting Methodology that is used for above-mentioned performance parameter, it uses following each parameter to calculate the estimated performance parameter value: the dissemination channel characteristic of the environment of network measure result or the value that is obtained by additive method, wireless frequency environment, the distribution of three-dimensional physical network, ad-hoc location and the specific physical distribution of assembly.

Summary of the invention

The present invention can predict, measures and optimize the performance of data communication network.The present invention can show the distribution details that disposes communication network complete or expection in the model of a physically accurate computer representation or three-dimensional environment.This makes the present invention use to be used to the customizing messages of creating network of network entity, assembly, subsystem and system actual or expection, can stores it, and in the ad-hoc location of physical environment is represented, determine performance prediction.Measurement Agent with known or three-dimensional position of distributing is used for measuring the performance parameter in the original place that is sent to processor-server.This processor-server has the accurate three-dimensional model of environment, can handle the data that record, and also can use independently or utilize the information of the ad-hoc location that records data, and forecast model is provided.Processor-server can be communicated by letter with other processor-servers in the mode of classification, make and to collect from many data long-range or network of location altogether, and be used for measurement result is shown and classifies that these measurement results may be used for or be not used in the model of creating estimated performance.In addition, each processor-server only consider simply that the ad-hoc location of physical assemblies distributes and physical network in specific delays time, transit time, communication effect, multipath and noise factor, can not use the data that record and calculate the model of estimated performance.

The present invention can predict throughput, bandwidth, service quality, bit error rate, packet-error-rate, the frame bit error rate, packet loss, packet awaits time, return interval, propagation delay, transmission delay, processing delay, queueing delay, network capacity, packet jitter, bandwidth-delay product and switching delay time in mode ad-hoc location, three-dimensional model in any environment.The present invention can measure and predict above-mentioned all properties standard, and the result is deposited in the accurate three-dimensional model physically in data communication network and the institute's installation environment thereof.And the present invention can show performance standard that record and the data predicted communication network in the mode three-dimensional under any environment, ad-hoc location.These performances provide strong design environment for wired and wireless network, it makes those of ordinary skills can be in mode three-dimensional, ad-hoc location, use unprecedented method, design apace and simply, measure, predict, optimization and visualized data communication performance standard.

Brief Description Of Drawings

Fig. 1: the example that transmits data by communication network

Fig. 2: create digital signal by simulating signal

Fig. 3: the difference between bit, grouping, the frame is described

Fig. 4: data shown in each node are described in the data communication network tree derivation

Fig. 5: the method for the ad-hoc location model of the three-dimensional of creation environment

Fig. 6: the method for using the prediction optimization data communication network

Fig. 7: use measurement result to optimize the method for data communication network

Fig. 8: use prediction and measurement result to optimize the method for data communication network

DETAILED DESCRIPTION OF THE PREFERRED

The invention provides the ability of design, measurement, prediction and optimization data communication network performance.The three-dimensional model that is stored in the communication network in the computer data lab environment accurately that the present invention uses a computer and generates.The present invention makes the user network cable, hub, router, switch, bridge, WAP, amplifier, shunt, antenna (some type, omni-directional, direction type, leak feedback type, profile, array type etc.), transceiver, terminal and other communication and computer network facility can be positioned over the modeled physical location of its reality.The present invention uses the height precise analytic model of the physical distribution of this structure, makes the user can be in modeled physical location any three-dimensional, ad-hoc location, and the performance of any communication network is carried out visual, prediction and optimized.

Embodiments of the invention can connect and the physical location aspect from logic, and the hardware of the communication network of ad-hoc location is carried out modeling.The present invention uses the logic Anknupfungskeregriff (being called as topology distribution sometimes) of known classification, and described notion is applicable to the data communication network with the model combination of a physically accurate ad-hoc location.Invention is in the past only paid close attention to the topology between the networking component or is concerned, distributes.The present invention uses specific three-dimensional model, thereby can provide accurate equipment control of the height of actual installation equipment and tracking, and the network performance prediction of the physical dimension that uses accurate network, the ability of measuring and designing are provided simultaneously.In addition, the peculiar and device specific performance catalogue of important network of all article that use in the present invention's while storage networking is as manufacturer, module No., network hardware type, operating system version, firmware and software type and version.The Network Layering tree shaped model is called as Site View, the attribute of selecting special clauses and subclauses in the three-dimensional environment model or node can show, store or print each device by it.And, using Layout View or Site View, reposition can be replaced, deletes, disposes or be moved to the network hardware and component software in real time alternately.Below will further describe in the model of the three-dimensional ad-hoc location of environment and to follow the tracks of with the accurate dimension of real space position and every kind of method of planned network, described method is used for the bill of materials of data communication network of model of creationization, and the patented claim of pending trial " is used for designing or disposing the method and system of the communication network of considering component property " and has described one of them preferred embodiment in application on August 4th, 2000.

Fig. 4 has represented the hierarchy of a data communication network, as an example of contained information among the Layout View.In the drawings, come all hardware in the display network with tree structure.Each node on the tree comprises the true physical location of tracking data communication network hardware, the information that the logic distribution is connected with electricity, light and electromagnetism of being used for, version and setting with software that moves on the network equipment and firmware, and the known performance parameter of the network equipment, comprise throughput, bandwidth, service quality, bit error rate, packet-error-rate, the frame bit error rate, packet loss, packet awaits time, return interval, propagation delay, transmission delay, processing delay, queueing delay, network capacity, packet jitter, bandwidth-delay product and the switching delay time of device.

Site View of the present invention has physically accurate three-dimensional modeling ability, with the all-network device in the model of the ad-hoc location of display network environment of living in.That is the preferred embodiments of the present invention make it possible to place each modeled hardware and software device with the accurate way of three-dimensional, and the attribute of the tracking device relevant with data communication network.These determinant attributes comprise following project: high and low and " standard " index of type of hardware, hardware configuration, software type, software arrangements, operating system version and each assembly.These indexs can comprise important device and network subsystem operating parameter, as throughput, bandwidth, service quality, bit error rate, packet-error-rate, the frame bit error rate, packet loss, packet awaits time, return interval, propagation delay, transmission delay, processing delay, queueing delay, network capacity, packet jitter, bandwidth-delay product and switching delay time.As described below, Site View has considered difficulty, and solved data network prediction, design and optimization problem in the more complicated data communication network, and replace for above-mentioned Wireless Valley Communications the prior art described in the patented claim of Inc pending trial.Especially, this new invention has considered to comprise modeling technique, performance prediction method and the method for designing of ad-hoc location of the physics of wired and wireless data network system, and wherein said data network system has the performance characteristic than the much complicated physical factor that only comprises wireless signal strength, interference and multipath.Especially, for data communication network, many and particular network device or other relevant factors of modulator-demodular unit design, as the multipath transmisstion effect of the size of dividing into groups, equalizer configuration, modulator approach, source and error code coding method, packet oriented protocol, the number of users of shared channel network, the lasting type that is used for transmitting again grouping, wireless system etc., the extra factor that must take in when the design of communications network is provided, and described communication network is used for data service rather than simple simon says sound business.

At present a difficulty facing of network designer or network system administrator is that majority of network equipment uses proprietary, private method to implement different network equipments, and the difference of the manufacturer of these methods Yin Teding and difference.Therefore, in wireless network, only use the basic physical propagation model to be difficult to set up reliable forecast model.Along with such as bluetooth, DSL, according to the voice of IP and increase in the future according to the data transmission technology of the cellular radio structure of grouping, prediction and the ability of measuring the performance parameter of particular network will become and become more and more important, and the ability that suitably measurement result is attached in the three-dimensional prediction model of performance parameter will be very important for suitable network configuration.

The present invention has considered the attribute relevant with the packet-switched data communication network, it uses the individual consumer's (or user of restricted number) of a special use traditional cell phone or voice call communication system to compare with using circuit switching and each assignment channel, and needing more extensively and not is conspicuous modeling.Data communication network has at grouping system specific performance properties standard, is not suitable for all types of communication networks of above-mentioned expection.Therefore, the preferred embodiments of the present invention can be according to the specific physics of each networking component and the physics of locus and specific components, electric and logical attribute, prediction throughput, bandwidth, service quality, bit error rate, packet-error-rate, the frame bit error rate, packet loss, packet awaits time, return interval, propagation delay, transmission delay, processing delay, queueing delay, network capacity, packet jitter, bandwidth-delay product and switching delay time.Performance prediction method uses the bill of materials of constructing in the environment of three-dimensional modelization, consider all devices and the network equipment, the interior physical location of environment that comprises three-dimensional modelization, and can carry out performance prediction to any desired position in modeled network and the environment, according to user's demand, described position may be in the ad-hoc location, building in the room, in the room or the outdoor area of different interval.

Can be by the performance prediction of all the cable network assemblies performance prediction with wireless module be separated, and then with result combinations, obtain total network performance, predict throughput, bandwidth, service quality, bit error rate, packet-error-rate, the frame bit error rate, packet loss, packet awaits time, return interval, propagation delay, transmission delay, processing delay, queueing delay, network capacity, packet jitter, bandwidth-delay product and switching delay time and other performance parameters.When predicting the performance of wire communication link, the effect that is provided with in conjunction with the known particular network of each wireline equipment, be operating parameter or performance parameter, be very important, these parameters comprise the typical case of protocol type, data type, grouping size, professional operating characteristic, firmware type, OS Type, representative network Performance Characteristics, network, average, peak value and minimal service load etc.For wireless network components, also must consider the extra factor of relevant propagation, signal intensity, interference and noise.

The preferred embodiments of the present invention make that with the method for a lot of novelties the performance of data communication network is predicted accurately.

At first, can detect according to the scene of a real network and carry out performance prediction, wherein from some cooperations of measurement data, form forecast model (empirical model).These in-site measurements result can manually carry out, and also can use data acquisition unit or the Agent of measuring and be updated in observed particular network performance standard in the physical environment continuously to obtain automatically.Specific three-dimensional localization position can be measured or distribute to these data acquisition units in physical environment, these positions are corresponding with the known locations position that is used for the network physical environment is carried out in the computer model of modeling, and the latter is known or is transferred into the measurement server.Data acquisition unit can be the independent device that manually or automatically writes down or gather the network performance of observed all one or more performance parameters as previously mentioned, measure Agent and can be in order to measure one or more above-mentioned network performance parameters of listing termly, and operate in software, hardware or the firmware application programs of network application top layer.Agent can be fixed, and also can be of portable form, and they can have locating device, and the internal map as GPS or inertial navigator or user's startup is sent to processor-server with the position location with measurement result.Suppose between Agent and processor-server to have two-way communication, this server can be set together the configuration or at far-end.By wireless or wired mode, or mode real-time or storage, collect from the measurement result of one or more data acquisition units termly or periodically, and it is sent to the processor-server that is set together with it or be located at far-end from one or more Agents.For example, can be by the record of Agent independently measurement result, and on fixed network, be sent to processor, comprehensive all measurement result and the computer statistics numerals of this processor are observed.The measurement source has known or unknown three-dimensional localization, and is used for forming the guestimate of the performance of observed network.Can be through transmitting in real time, store and transmit such as the Email, the different modes wireless, wired or optical link that on WWW, transmit with file, the measurement result of collection is perhaps in memory storage.The server that this " scene " measurement data transfers to classification and handles particular measurement information with available three-dimensional localization position.Use is from the metrical information of data acquisition unit, the knowledge of the three-dimensional environment by using physics, and many data that collect are dissolved into the performance model relevant with the three-dimensional physical representation in the world of simplification, server can provide forecast model.

In a preferred embodiment of the invention, server stores is also handled the physical location of all (available) measurement mechanisms, and all-network assembly and electric, logic and technical configuration, also considers expense and the maintenance issues relevant with each networking component.Use the preferred embodiment, can be by collecting from the measurement data of one or more Agents, and server process they to determine that before installation one is the prediction engine that network topology in future provides desired effects, comes the data communication network is designed, disposes, tests, predicts, measures, optimizes and safeguards.Server engine can show that each measures the measurement result of Agent from (comprising specific location information) in the mode of ad-hoc location, like this can be on the display device of computing machine or memory storage comparison prediction and measurement result (such as the ascii text file of a performance parameter comparison prediction and measurement).

It should be noted that each measures Agent can be a server, this server can be fused to measurement data in the three-dimensional layout of ad-hoc location of networking component and physical environment.Therefore, each is measured Agent and also can be used as central processing unit, thereby can measure and predict the parameter of many different physical locations of particular network.Server can be set together with the measurement Agent of gathering, show, storing and use measurement result to form forecast model or be positioned at far-end.Being positioned under the situation of far-end at the server that receives from the measurement data of measuring Agent, is possible with the far apart network performance of particular server processor physically from remote monitoring and prediction.

Processor-server can further be controlled or the allocating and measuring Agent, to adjust or to indicate Agent to finish dissimilar measurements, as different packet transmission rate, observe interval, equispaced, protocol type, or those of ordinary skills think other significant changes that are used for the suitable network optimization.

The method of second kind of prediction network parameter performance is the method by operational analysis or simulation.The method of these analyses and simulation is well-known, and makes the physics of network channel relevant with the physics and the electrical specification of variety of network components with electrical specification.By simulation and analysis, can determine the approximate or border of desired representative value in the real network configuration of specific components.Embodiments of the invention allow the user to import these result calculated, thereby make their inputs as forecast model.Therefore, user of the present invention can import " blind " value simply according to known method, as the initial guess that forms the network performance forecast model approximate (first guess approach).The present invention's feedback that can obtain according to the measurement result from the ad-hoc location of real network is carried out iteration to these preliminary examination guess values then.

About measuring the description of Agent and processor-server, the present invention writes down, stores and be presented at a series of measurement data of the typical operation environment that has a plurality of senders in wireless or the cable network as above-mentioned.Then certain optimal algorithm (lowest mean square, medium filtering etc.) is applied to the model of the prediction that from following equation, obtains, so that the suitable performance number that is used for the determining specific ad-hoc location network design question blank of (as the constant in the following performance parameter equation or the desired value of function) to be provided.This look-up-table method makes the data that record be translated into the value that is used to drive predicted data subsequently, and this predicted data is used for all subsequent prediction of carrying out in the three-dimensional environment of the identical ad-hoc location of testing.In addition, network environment for specific three-dimensional modelization, the performance prediction of any space of three-dimensional environment or position is according to initial, the best forecast model of inferring in this environment, via memory storage or from wireless or non-wireless means far-end or that be set together, the best is inferred the function in performance standard value or the following equation or the best guess value of constant, automatically or manually feed back to the present invention.The estimated performance of any locus of three-dimensional environment all is according to the initial best forecast model of inferring in this network environment.As explained below, these are initial best infer or " blind " model can be according to simulation, analyze or their certain combination.The experience forecast model with initial best infer forecast model can be used for subsequently with measure or the best different environment of inferring of environment, and invention makes the user can be easily with model classification, to be used for network prediction or design subsequently.Then, the measurement result of real network performance can side by side cover, show and store with the network Prediction Parameters, to compare fast.And, the optimum value with least error of the new forecast model that the network performance that records in optimizer calculating and the environment is complementary.Therefore, the present invention allows the user to use the customizing messages and the physical location of network equipment, and the model that uses cable network, radio transmission, multipath and noise, in the structure of three-dimensional ad-hoc location of network actual installation or expection, be related with the estimated performance parameter of experience gained or the network performance parameter of initial guess.Modelling technique of the present invention merges many factors that influence network performance in better simply model, and this model supports is to the comparison between the network performance of the prediction of wireless and cable network and the network performance that records.Therefore, can determine performance prediction, and itself and the network performance that records are compared, in network configuration, to use.

And, by the network performance index that relatively records and the index of prediction, the present invention makes new in-site measurement result can upgrade the forecast model of front with a kind of method easily, this easily method provide by the Local or Remote storage and be shown to user's model classification.In addition, by using the classification of server, might make that calculating, transmission that is positioned at far-end or the server that receives this measurement data and forecast model are used for remotely using, show, measurement and stored parameter model and result.This for hope concerning physically with the far apart network of interested network carry out the network manager of network performance and design monitoring convenient especially.

Can the preferred Survey Software module of the application of the invention, or from another software or hardware measurement tool, import log file and carry out measurement to the specific device of desired properties standard.The preferred measurement module of the present invention uses the two or more software software programs of installation and operation on device, can measure the performance of arbitrary specific part in the communication network.These software programs are called Agent.By sending the test transmission between two Agents that connect in particular network, the present invention preferably can measure any specific performance properties standard.These measurement results are stored to be used for the specific part of network.

The preferred embodiments of the present invention can also import the log file such as other processs of measurement of Traceroute, to measure specific link.The external program that this function allows to store in the ad-hoc location mode carries out the measurement of ad-hoc location.By realizing this function by a kind of bilateral method of the patent 09/221,985 of the T.Rappaport and R.Skidmore application description that " is used for creating the computer model of wireless communication system and the system of measured database " on Dec 29th, 1998.In order to import log file, for each data point, the user clicks a bit in the environmental model simply, and the every bit that just can be in the log file distributes a position.

When carrying out the performance measurement of network, especially for radio data network, the performance difference of understanding between territory and the reception is very important.Here it is, and why the present invention preferably can measure the transmission of averaging network statistics and the reason of receiving unit.For measuring transmission direction, need to change the size of test packet.By grouping that change to send and the size of returning grouping, can separated transmission and the statistics of reception.This makes network designer can be clearly may be covered problem in the transmission by reception good on the surface.

If measure the actual data traffic of carrying on can not analog network, then network performance measurement is useless.Therefore, the preferred embodiments of the present invention can be simulated business model, procotol and the packet characteristic of real data.Therefore, if test network is browsed performance, then the present invention sends little grouping from accessing terminal to the web server, and returns the big grouping of text, image and Networking script file layout typically from server.By measuring the performance of these groupings, the present invention has accumulated the network statistical data that is used to the network browsing performance expected accurately.

Also the measurement of particular traffic type can be applied to use the situation of broadcasting or multileaving packet performance.The preferred embodiments of the present invention can be measured a plurality of senders or the sender of a plurality of receiver or same packets information and the performance of receiver.The performance of this transport-type is different with the measurement of point-to-point, because in the situation of broadcasting and multileaving, more effectively utilized shared resource.Therefore, measurement of the present invention is very powerful about the ability of the network performance statistics of wireless and multileaving and sender that each is independent and receiver.This function makes network designer can select better which sender is redundant in the multileaving, and perhaps which wireless transmission is not enough to reach all required receivers.

In some data communication networks, measure such as Ethernet bridge or even the performance of the particular device of a cable be difficult, reason is it for the network layer of data communication network is transparent.Therefore, the present invention determines that by extrapolation method the function of the performance of single assembly is of great use.The preferred embodiments of the present invention can be used the performance data of known network specific device, and the effect of other devices in the extrapolation network.By measuring and extrapolate abundant independent hardware and software link, can determine the performance of all-network device.The accuracy and reliability of this process depends primarily on the model of the accurate ad-hoc location of the data communication network that the present invention has.

With the extrapolation concept extension of performance evaluation software and hardware assembly, shown the further function of the preferred embodiments of the present invention to the network equipment.In some cases, the present invention can distinguish because the caused performance limitations of software and hardware.For example, be under the situation of same computing machine at sender and receiver, do not relate to hardware in the actual transmission.In this case, by the Measurement Network statistics, can only evaluate the performance of computer software.Is the situation of various computing machine by comparing sender with sending identical situation and sender and receiver, can only determine the performance of computer hardware.Since in two kinds of situations, because the performance of software is quite similar, therefore can be similar by supposition software performance in two kinds of situations, infer the performance of the hardware that connects two computing machines.

Process by the performance of the performance index that record extrapolation stand alone network component may be more time-consuming.Therefore, the preferred embodiments of the present invention can be read in the data result from many survey instruments, system program and Network records file and are simple internal format.The present invention can be read in such as the output of the order line program of ping and ttcp, such as the log file that is produced by router and switch of Tcpdump or the log file of other commercial processs of measurement, and, use for prediction engine with these measurement results preservations.By the combination of files with these importings is simple internal format, and the present invention can be combined as a simple network statistical data collection with many different measuring results and activation record.This process means less activity measuring phases of the present invention's needs and more different, accurate data, comes better and more carries out the network performance modeling.

Data point that need to represent a large amount of measurements accurately, reliably to the data communication network.Therefore, the preferred embodiments of the present invention are used above-mentioned the whole bag of tricks, gather mass data apace, easily.The present invention accomplishes this point by long-range data collection agent program is provided, and this Agent can be installed on the data access terminal, or is embedded in hardware, software and the firmware of network actual device.The remote data acquisition Agent is in response to server program (processing server), and this server program is controlled the measurement result that remote agent program obtains.That is, can indicate the remote agent program to use any required agreement, type of service, network settings or configuration, arbitrary other remote agent program or processing server measured.This process does not need the user to carry out any input at the physical location of remote agent program.When requiring correctly to be provided with, Agent is record data simply, and the result is reported from the server of the data of all remote agent programs and other survey instruments to storage.Server can produce various detailed reports, and uses the network performance in future of data prediction expection.Server is as Agent.Under this mode, can constitute server in mode level or distributed.This makes between the server can report measurement result each other, and uses other Agents or server to measure.Therefore, the network designer of server end can be used that all collect and data report, come to determine such as the rationality of broadcast data or the problematic zone of bad distribution, or problem time arranged such as what the network activity of dinner hour data communication network increased.

For the value of the measurement data that improves collection, the preferred embodiments of the present invention are determined accurate (if possible) or the apparent position of remote agent program.As discussed above, in this case, can be by the user of physical location control or by server control agent program remotely.In a preferred embodiment of the invention, Agent uses the information of network topology to determine an apparent position.By definite nearest network equipment, and the position is associated with accurate, the known position of this network equipment nearly, realizes this function.Can further use dead reckoning, click the position on the map or use GPS, laser range finder or other now or in the future known devices come this apparent position of refinement.

The preferred embodiments of the present invention not only can be assessed the influence that different hardware, firmware, software and configuration are provided with, and only can also predict by hardware and firmware, only by software or the influence of generation only is set by an independent configuration.The measurement that the present invention carries out these influences reaches and then the ability of adjusting experience prediction makes and can be optimized the data communication network.Change the data communication network caused influence in arbitrary detailed aspect by prediction, the user can be immediately with New Parent or that the influence of change is set is visual.This ability makes the user of ability field those of ordinary skill design the data communication network of an optimization by making the change of changing and observing prediction continuously.

Now according to the specific components of using in the environment knowledge of the ad-hoc location of three-dimensional and the network design, we will pay close attention to the detail of predicting the network performance parameter value.

According to the present invention, the throughput and the bandwidth of coming computational grid with the function of any one or all parameter in the following operating parameter that influences performance: the distance between sender and receiver, physical environment index, grouping size, error code and source code mode, packet overhead, modulation technique, environment, interference, signal intensity, number of users also have antenna mode and type, multidiameter, the quantity of multipath component, arrival angle, radio frequency bandwidth, agreement, coded system and the three-dimensional position of multipath component for wireless network.For bandwidth and the throughput of predicting that network connects, can from listed parameter, calculate suitable function and constant, predict for each position and required time then.

For cable network, can on the product or device instructions of manufacturer, perhaps in being provided with, special laboratory measure throughput (T) or bandwidth (BW).Under another situation, can pass through to analyze or analog computation, or use many known devices to measure T or BW at the scene.These methods can be used for determining such as T in the network prediction engine discussed herein or the desired value of BW.Formula 1 is the formula that is used for predicting the throughput and the bandwidth of wireless data communications channel.

Wherein T is a throughput, and BW is a bandwidth, and d is the distance between sender and the receiver.RSSI is the signal intensity density that receives, and promptly the power level of receiver end signal can be absolute value or logarithm value.A, B, C, C ₁, C ₂, C ₃, D, E, F, K _IBe constant, perhaps can represent such as physical environment type, grouping size, modulation, modem type, or the one or more physics of the parameter of other and physics, electric or cellular logic environmental correclation or the linearity or the nonlinear function of electric parameter.Whether according to the T or the BW that are separating, these constants or function are got specific functional value.Value M can represent the specific quantity from the multipath component of specific sender, and it is analyzed by the propagation of channel and determines; M can also represent the combination from the important multipath component of a plurality of senders, and " important " is to depend on that antenna mode, physical environment distance and other those of ordinary skills know and the following radio transmission factor that will explain: G here _iAnd P _iValue represent the gain and the power level of each signal in the different component of signal of M respectively, they may represent one multipath component or from total component of signal of one or more radiation sources.K _iRepresent a Finite Number of the constant or the function of each i value.Note G _i, P _iWith each K _iCan be logarithm (as dB) or absolute value.These constants in the top equation or function can depend on the distance (d) of sender and receiver, wherein d can be the reflection/diffraction distance of main signal between air line distance or sender that is moving and the receiver, also depend on three-dimensional environment, observing time or observation are at interval, noise power, the grouping size, coded system, number of users, modulation type, disturb, for wireless network, also comprise path loss, multidiameter, the multipath component number, angle spread, the intensity of received signal and arrival angle, modulation band-width, and other physics of special equipment in the network, electric and logic setting can be calculated with resolving, prediction preliminary examination value is inferred, perhaps use the method for the optimum matching between the performance of the performance of the measurement in the real network of ad-hoc location environment and prediction to calculate these constants and function.

It should be noted that as described herein and as equation (1) shown in, can consider in many ways that multidiameter and it are to network performance prediction and the influence that designs.At first, can consider a plurality of paths independently, each multipath component regarded as by each dispensing device arrive, in the method prior art of multipath modeling and in many research work of the people such as Rappaport of Virginia Institute of Technology, be known and obtain explaining.In addition, the footpath influence of total pen is modeled as the delay (as propagation distance d) that has under the least favorable situation, is similar to by maximal value, mean value and intermediate value and obtains the length of the three-dimensional environment of communication network modeling and specific building.In addition, can be by the antenna mode of expection each sender and receiver, usage space considers, thereby during the network performance in computing relay and (1), only considers the multipath situation that the main beam at each wireless device arrives.In addition, the multipath transmisstion that can only consider the strongest one or two or a limited number of sender postpones, thereby only consider a finite set of sender, as distance receiver sender nearest or that have certain standard, frequency or power setting, launch multipath energy and produce the RSSI value, when computing relay, only consider the strongest multipath or mean value, maximal value, intermediate value or maximum several multipath components at these a limited number of senders.In addition, if only consider a limited number of sender, then such as by considering that physical environment determines from total multidiameter of each sender or use specific antenna mode to determine the said method of most important multipath component, can be used for setting up the model of multipath and it influence to network performance.Can come the signal intensity that receives, the RSSI modeling in the equation 1 with similar method.

Notice that constant or function in the equation (1) can at random distribute the initial predicted value, measure the particular network in the environment of ad-hoc location then by rule of thumb, thereby can distribute the constant in (1) or the value of function with optimum matching (using Minimum Mean Square Error or other known methods).Note, in (1), distance (d) can be the actual physical distance that obtains according to the ad-hoc location model by the three-dimensional of environment, also can actual represent a relative distance rates, wherein the physical distance of point-to-point transmission is sealed free space reference range (close-infree space reference distance) easily with reference to one, and it is generally used for propagation forecast, at Rappaport, " Wireless Communications, Principle﹠amp; Practice, Prentice-Hall introduces in 1996 to some extent.

The propagation delay of network data is the prediction cable network, assembly interconnects by electric wire (optical fiber or metal cords) in cable network, by with the velocity of propagation of propagation distance, draw propagation delay divided by electric, electromagnetism in the device that is used for transmitting data or light signal.For example, because the non-transport properties of cable has influence on the photon of the fiber optic cables that are used for transmitting data, so the speed of data is 2*10 in the optical cable ⁸Meter per second.This photon moves in glass with the light velocity, is lower than the velocity of propagation of free space.Therefore, if optical cable is grown 200 meters, then transmission delay equals 1*10 ⁶Second.Use the method for the ad-hoc location to the complete network modeling of the present invention, the user can side by side carry out network visual as disposing in the environment, and sees delay and the prediction of delay or the demonstration that records performance in the cable in the three-dimensional environment.In addition, use mouse or other indicating devices, perhaps use drop-down menu, perhaps simply by observing display device of the invention process, can be visual or storage diverse network performance index and bill of materials in the data of storing and the parameter of interpolar.

Because a plurality of senders source such as the access point in blueteeth network, IEEE802.11b or the Wireless ATM Network may side by side send, so the propagation delay of the wireless portion of predicted data communication network is more much more difficult than cable network.And as mentioned above, multipath disturbs can produce echo, and can echo be depended on wireless receiver or the employed particular network device of sender by equilibrium.But, can use the computation model identical, and, consider multidiameter and propagation loss or gain extraly, shown in equation (1) by the multipath component generation according to specific multipath component with being used for cable network.Because wireless data is always with rectilinear propagation, and object can diffraction, reflection, absorption and scattering wireless energy, so need consider multidiameter extraly.Therefore, for the transmission delay of Radio Link in the computational data communication network, the distance between sender and the receiver is divided by the velocity of propagation (3*10 of wireless communication link ⁸M/s), add the multidiameter of introducing to non-directional route the receiver by sender, shown in equation 2.

$T_p=\frac{d}{3*{10}^{8}m/s}+{\mathrm{\τ}}_d---\left(2\right)$

T wherein _pBe propagation delay, unit is second; D is the distance between sender and the receiver, and unit is a rice; τ _dBe multidiameter, unit is second.Use known ray trace technology or, come multidiameter is predicted according to arriving angle or signal strength values or making an estimate according to the physical model of three-dimensional environment.

Use the bit number that transmits, can directly calculate transmission delay by the bandwidth meter that connects.For calculated transmission delay, with the bit number that transmits divided by bandwidth.This calculating for wired be identical with wireless channel, but for the computing alone of each network equipment.Formula is shown in equation 3.

$T_t=\frac{\#\mathrm{ofbits}}{\mathrm{BW}}---\left(3\right)$

T wherein _tBe the propagation delay time, unit is second, and #of bits is the bit number in transmission or the grouping, and BW is the bandwidth of network link, and unit is a bits per second.

Must postpone each device computing in the network individually.Processing delay is that network equipment is used for handling, storing and transmit the required time of data bit that is applied to network equipment.In addition, processing delay can be the source after obtaining indication, produces a needed time of significant data stream.For the device that does not carry out any processing such as the passive network assembly of cable, antenna or shunt, its processing delay is 0.Processing time is depended on the current computational load of grouping size, protocol type, operating system, manufacturer, firmware, hardware and software version or configuration and type of device and device.For predicting the processing delay of any device, must use a model of considering all these influences.These models can record at the scene, record in test procedure, by obtaining from manufacturer or by analyzing or simulation obtains.

Queueing delay is only applicable to from the device of multi-user or multi-link transmission data.The queueing delay of device is that specific cluster must be waited for the required time quantum of other professional transmissions.Because it depends on the handled portfolio of specific device, is difficult so predict the queueing delay of specific connection.Therefore, the statistics stochastic variable of the estimated performance of operative installations and/or expection business load can be predicted queueing delay.In addition, by the device technique index definition or measure, simulation, the mean value that calculates the queueing delay time of gained, intermediate value, preferably or worst condition or some other linearities or nonlinear weight, can be used for calculating prediction queueing delay time.

Packet awaits time, return interval and switching delay time are all depended on propagation, transmission, processing and queueing delay time.In order to calculate to a nicety packet awaits time and return interval, propagation, transmission, processing and the queueing delay time of all-network device in the particular network link must to be added up, and use specific type of service, grouping size and protocol type to adjust.For example, the packet awaits time is to divide into groups from sender to the receiver required time.In order to predict the packet awaits time of particular link, must use the size of the grouping of particular network connection, type of service and one way transmission, calculate propagation, transmission, processing and queueing delay time.

The calculating of return interval is herewith similar, just comprises grouping and sends and receive, and confirm returning of grouping.Therefore, for predicting return interval, the present invention considers the influence to propagation, transmission, processing and queueing delay of the size of original packet, the size of confirming grouping and particular network connection, agreement and type of service.

The switching delay time depend on relate to two independently propagation, transmission, processing and the queueing delay of WAP have, described WAP is coordinated the change of wireless device from an access point to the control of another access point to therebetween.The generation of these delays is because two access points must transmit data back and forth, to complete successfully switching.Therefore, the prediction of the packet awaits time between the prediction of switching delay time and two access points is similar.Be the prediction switching delay time, the present invention calculates propagation, transmission, processing and the queueing delay of the link between two access points.The present invention adjusts the specific quantity of required transmission then, and for successfully switching the size of the data that must send.

When the prediction bit error rate, the present invention considers the wired and wireless bit error rate.Compare with cable network, wireless network complexity move in the many electric circumstance, they interconnected more is difficult to modeling, and, before the present invention, also not in the environment of the ad-hoc location of three-dimensional, use specific, the accurate physics and the electronic model of a plurality of senders, a plurality of interference source, noise source and networking component, successfully to the real network modeling.The present invention uses the three-dimensional ad-hoc location of the environment of particular network enforcement to represent, it has considered wired and wireless network in the three-dimensional environment model of real world, and considered the physical location of all radiation sources and antenna system thereof, electric index and characteristic, these network equipments can be regarded wired and wireless network as.As previously mentioned, because multipath transmisstion, multipath transmission and The noise make wireless network than the easier generation error in data of cable network.Must consider that when actual design radio transmission and noise ratio are fixed with spider lines more at random, this is modeling in the present invention.For wire message way, bit error rate is exactly the measurement of connect electric, light and electromagnetic parameter, can use such as Gauss or Poisson stochastic distribution or other the known tangible distributions that maybe will know or the statistics stochastic variable of algorithm and predict, this stochastic variable has comprised average, the intermediate value or the typical performance of networking component or network subsystem.Network equipment or subsystem can comprise independent radio node, as router or switch or by different routers, hub, switch, WAP and with the radio customer machine/server unit of network service form complete interconnected.Network can the time wireless, wired or the two combination.

Can be from the technical indicator of device auto levelizer or network subsystem such as the frame bit error rate, bit error rate, or many performance index of packet-error-rate, and such as throughput, bandwidth, service quality, bit error rate, packet-error-rate, the frame bit error rate, packet loss, the packet awaits time, return interval, propagation delay, transmission delay, processing delay, queueing delay, network capacity, packet jitter, other performance parameters of bandwidth-delay product and switching delay time, also can use the present invention to calculate or with its input the present invention with analyzing, or as using the priori value of measuring before the present invention.That is, can predict special parameter such as above-mentioned listed operation, i.e. operating parameter or device parameter by the technical indicator of the device measuring or provide by manufacturer.In addition, can measure them, with the modeling that is used to be fit to and import the present invention by user or research tool on-the-spotly.In addition, can calculate them according to some interconnected analytical model of device, thus can calculated performance model and operating parameter.Among the present invention to the required statistics stochastic variable of network performance modeling can depend on each device such as voltage level, power level, impedance, and electric, light and the electromagnetic property of operating frequency, perhaps can use typical case to observe the generation of (measurements) value to each network equipment.For example, copper cash can be modeled as bit error rate is every transmission 10 ⁶Or 10 ⁷Bit produces an error code.In case measure and to after the independent initial time characterization, the present invention can repeat the independent assembly in the network or a string assembly modeling, thereby to the network performance modeling.

But, to compare with cable network, the bit error rate parameter of wireless network depends on more factors.Therefore, the present invention predicts bit error rate according to following factor: number and type, time, three-dimensional position, grouping size, protocol type, modulation, wireless frequency, wireless frequency bandwidth, coding method, error code correction technology, multipath signal strength, arrival angle and the multidiameter of the distance between environment, sender and the receiver, transmission interruption.As a result, use constant or function that the bit error rate of prediction is calculated,, be converted to the bit error rate of expection that record previously or known channel and the performance index network equipment.Equation 4 is to be used for the directly formula of bit error rate, the frame bit error rate or the packet-error-rate of predicted data communication channel, and it is identical with equation 1.

Wherein BER is a bit error rate, and FER is the frame bit error rate, and PER is a packet-error-rate, and d is the distance between sender and the receiver.RSSI is the signal intensity density that receives, and wherein intensity density is the power level of receiver end signal.A, B, C, C ₁, C ₂, C ₃, D, E, F, K _iBe constant, perhaps have the linearity and the nonlinear function of different value, wherein this value depends on and which calculates among BER, FER, the PER and decides.The M value can be represented the specific quantity from the multipath component of specific sender, M also can represent the specific quantity from the combination of the important multipath component of a plurality of senders, and " important " is according to antenna mode, physical environment distance and those of ordinary skills are that know and other radio transmission factors that obtain in the present invention explaining here.M G _iAnd P _IEach value in the value is represented the gain and the power level of unlike signal component respectively, and they may be represented from the independent multipath component of one or more radiation sources or total component of signal, can be the logarithmic form (as dB) or the linear value of power.Variable G _i, P _iWith each K _iValue can be logarithmic form (as dB) or absolute value.These constants in the above-mentioned equation or function depend on the distance (d) of sender and receiver, and wherein d may be the sender of work and the air line distance or the reflection/diffraction distance of the main signal path between receiver.As the explanation to equation (1), this distance can be an air line distance, or can be obtained by the total characteristic modeling such as the environment of maximal value, average or the intermediate length of three-dimensional environment.Same equation (1) is the same, and equation (4) is regarded distance d as the actual physics distance, perhaps to sealing the relative distance of object of reference.

Can calculate the frame bit error rate, packet-error-rate and packet loss by bit error rate, or use method or the equation 4 identical with above-mentioned prediction bit error rate predicted directly.The present invention uses in the bill of materials be stored in ad-hoc location the information about the size of dividing into groups, frame sign and use agreement, and use is carried out these calculating such as the propagation and the interference modeling technology of the ad-hoc location that uses among the SitePlanner that creates in Wireless Valley Communications company limited.

In wireless network, it is extremely difficult that modeling is carried out in the influence of the combination results in all error code sources.Not only modulation and specific error code and source coding technique influence wireless network performance, and the specific RF compartment system that antenna, multipath, noise, IP-based voice or wireless ATM Cascading Methods, specific wireless modem manufacturer's specific modulator-demodular unit designed, was used to connect wired and wireless network all exerts an influence.Can carry out preliminary surveying and drive test, make the mobile receiver operation by it, and in the environment of ad-hoc location, carry out the measurement of network performance parameter, then, applicable equations 1 described notion might be determined the optimum matching of particular modem manufacturer.

In the present invention, can be by using directly computation bandwidth lagged product of following information: number, the intensity of arbitrary or all environment, three-dimensional location, protocol type, multidiameter, grouping size, wireless frequency, wireless frequency bandwidth, coding, multipath component and arrive angle, signal intensity, propagation, transmission, processing and queueing delay, bit error rate, packet-error-rate and the frame bit error rate.The present invention can also use the value of predicting previously, indirectly the computation bandwidth lagged product.The total delay that bandwidth and this device of particular network device causes is multiplied each other, promptly obtain bandwidth-delay product.Therefore, formula is represented by equation 5:

$\mathrm{BWD}=\frac{\mathrm{BW}}{T_{\mathrm{net}}}---\left(5\right)$

Wherein BWD is a bandwidth-delay product, and BW is a bandwidth, T _NetIt is the total delay of introducing.

The present invention uses the conforming statistical model of data communication network hardware to predict packet jitter and service quality (QoS).The index standard that these two performance standards all are network reliabilities provides consistent data to arrive number of times.Therefore QoS and the shake in order to calculate connection, the present invention uses the formula that comprises following information: quantity, the intensity of arbitrary or all environment, three-dimensional position, protocol type, multidiameter, grouping size, wireless frequency, wireless frequency bandwidth, coding, multipath component and arrive angle, signal intensity, propagation, transmission, processing and queueing delay, bit error rate, packet-error-rate, the frame bit error rate, throughput, bandwidth and bandwidth-delay product.Formula comprises constant or function, and they usually are associated above-mentioned variable with the variation of data time of arrival, and the QoS and the packet jitter that also above-mentioned variable can be connected with this is associated especially.Embodiments of the invention use equation (1) or (4) to come the QoS and the packet jitter of specified data communication network.

The influence of ad-hoc location, place layout, building layout, topographic(al) feature and RF characteristic has not only been considered in prediction according to a preferred embodiment of the invention, has also considered the influence of particular network hardware, firmware and software in the network.The bill of materials that the present invention uses the front to discuss makes and can carry out modeling to the setting and the position of each data communication equipment (DCE) of network.Considered also among the present invention that the prediction of network performance statistics also takes these settings into account.This means the agreement (as TCP or UDP) of different transmitting stages, different agreement setting (as grouping and buffering size), data bandwidth (is unit with the bits per second), comprise modulation technique (as QPSK or FHSS) Physical layer transfer approach, coded system (as CCK or grid coding), transmission medium (as copper, optical fiber or wireless connections) and special frequency channel.Consider position and wireless specific criteria in addition, comprise sender-receiver apart from the intensity of number of users, multidiameter, multipath component number and their arrival of (T-R distance), communication environments, interference, path loss, shared RF resource and the ratio that angle, related power account for uncorrelated power, and RF bandwidth (Hz of unit).The result that all these variablees obtain can be mapped in the form of equation (1) or (4).

Data communication network user's characteristic has been considered in the prediction of the preferred embodiments of the present invention.The present invention will store following information in the mode of ad-hoc location: the user is provided to type, number of users, the use pattern of the data communication service of network.That is, can place the point that some represent business that each user or this user provide or the zone of wherein distributing one group of user personality.The present invention will consider the zone of these points and customer service when the network performance standard is predicted.This means, if in the zone that access point covers discovery a large number of users, then the present invention's performance of these (or other) conditions that can calculate to a nicety.Exactly because the model of accurate, the ad-hoc location of data communication network just makes it become possibility.In addition, because the preferred embodiments of the present invention use pattern of constantly following the tracks of the user,, set up the constant in equation (1) or (4) or the question blank of function so processor-server can use measurement result.Can draw the constant in equation (1) or (4) or the different value of function, predict different network performances constantly in a day.Because actual network has peak service time and low level and smooth phase of use amount, so this is an important aspect for the data communication network forecast model.By the operating position of tracking data communication network constantly, the preferred embodiments of the present invention can determine whether have any problem at network sometime.

In communication network, the ratio form of the normally theoretical maximum possible capacity of capacity.Different user, their propagation characteristic, message size, and the influence of network characteristic has all limited the capacity that each user sees on network.Consider that a network has specific components or the device as bottleneck, its maximum rate is T _MaxBits per second.This component limitations the possible throughput of maximum of network.This capacity is represented capacity or throughput (being defined as T or capacity), wherein T (x, the y of device or network, z, t)=Tmaxp[γ], γ is a scale factor here, it is blended in a plian value between 0 and 1 with the physics of many different complexity, electric and logical condition.When γ is 0, there is not capacity.When γ is 1, have max cap..Notice that T is the function of three-dimensional position in the network, also is the function of time.For specific user, the target of network forecast model is that prediction is as the function of three-dimensional position and the capacity of the function of time.Therefore T[x, y, z, t] scope be 0 at T _MaxBetween.

Be added in each user's of load effect of data communication network capacity.Number of users and each user's use pattern influences the capacity of each user in the data communication network.The preferred embodiments of the present invention are measured instantaneous professional qualification by above-mentioned measurement Agent, make network designer can see the influence of offered load for the critical network statistics.Can be by determining on-the-spot cubic content measurement result such as observing the additive method that is built in the report mechanism hardware or the software product from the network equipment.Be generated as position and the function of the time in one day into the ad-hoc location of three-dimensional by the question blank with the specified vol result, the present invention has set up the capacity predict model based on measurement result.These measurement results are used to set up the model of capacity of the present invention.

Among the present invention, the scale factor of capacity (or throughput) is the function of following parameter: the average and maximum grouping size that instantaneous number of users, the maximum number of user of network while, the network user uses in the network, and modulator-demodular unit or network or manufacturer or specific many other factors of agreement.And for wireless network, propagation distance between multipath transmisstion influence, user and the WAP and received signal level all are the factors of limit capacity.In addition, also use the constant and the function of the influence of having merged modulation, equilibrium, impulsive noise and other factors among the present invention.

Therefore, the capacity of network and throughput can be passed through the following formula modeling:

Wherein the constant of (6) or function have and the middle similar characteristic of describing in equation (1) and (4).In addition, whole equation (6) can multiply by scale factor K/Umax, the instantaneous number of users of K formula network wherein, user's maximum quantity when Umax expresses possibility.

In wireless data communication network, the switching delay time is potential problem.In radio data network, when the user shifts out the scope of an access point and enters the scope of another access point, switch.In this case, first access point must be handed to second access point with the responsibility that transfers data to the wireless user.If two access points are at a distance of too far away, then the radio data network user will not have the enough time to switch to another access point from an access point, so file transmits failure.The present invention uses the measurement result of the ad-hoc location that depends on the time, and makes them be suitable for equation (1), (4) or (6), predicts where will switch, and because the incompatible possibility that causes handoff failure of network settings at two access point places.Then, can use table querying method determine the forecast model of switching time, this model is the function of locus and time.

The notion of optimizing is a key of the present invention.The preferred embodiments of the present invention are in the predicted value that makes those of ordinary skills obtain and store by the performance parameter that relatively records and the present invention, and the aspect that improves the performance of data communication network rapidly is efficiently.The process of using measurement result to improve prediction is called optimization, and described in Fig. 6, Fig. 7 and Fig. 8.Fig. 6 shows the method 6 of only using measurement result to optimize network, and Fig. 7 shows and only uses predicted data, Fig. 8 to show the combination of using measurement result and predicting the outcome.By a large amount of, visual or additive method, the position of more various ad-hoc locations and the process that the predicted value and the measured value such as following performance standard of special time is optimized data communication network in a day: throughput, bandwidth, service quality, bit error rate, packet-error-rate, the frame bit error rate, packet loss, packet awaits time, return interval, propagation delay, transmission delay, processing delay, queueing delay, network capacity, packet jitter, bandwidth-delay product and switching delay time.By changing the hardware that uses in the network, perhaps change the position or the configuration of the hardware of those hardware, firmware or softwares of in network, controlling each device, those of ordinary skills can improve network performance.In the three-dimensional model of network, behind the equipment of the ad-hoc location of change network,, can implement and observe these improvement in performance by repeating the prediction of performance standard.Continue this process, those of ordinary skills can optimize the performance of network, realize the active data communication network.

Use this information, the preferred embodiments of the present invention can and reconfigure for the upgrading of network suggested design is provided.The present invention can also use snmp protocol communication or other agreements, implements these changes practically.That is, Network Design person can be by predicting the problem in the specified data communication network, wherein use known measurement data and the configuration and the estimated performance of the hardware of all data communication in the data communication network, the prediction of performance standard that can the computational data communication network.Environment storage and show the performance standard of prediction visually with digitizing ground in the three-dimensional model of allocation really.Then, the deviser can use the present invention, by the change of recommending according to the present invention, perhaps makes the change of oneself, determines that those are by the predict the outcome solution of the problem that manifests of observation.Behind the simulation and forecast result, network designer can be indicated the present invention, and according to the change that the deviser has just used in prediction, all of updating apparatus are relevant to be provided with.The deviser can use measurement features of the present invention then, utilizes instrument to measure the result of these changes.

After having described the present invention by preferred embodiment, will recognize the present invention within the scope of the appended claims, multiple version will be arranged in ordinary skill.

Claims (84)

1. method of analyzing and adjusting cordless communication network may further comprise the steps:

Produce or use the computerized model of the cordless communication network in the physical space with computing machine or server, described cordless communication network is deployed in the described physical space, described computerized model provides the place layout, building layout, the specific expression in one or more positions in topographic(al) feature or the radio frequency features, the position of one or more assemblies in described physical space of using in the described cordless communication network of described computerized Model Identification, described computerized model has the modeled attribute of at least one assembly in described one or more assembly;

At described computing machine or server, receive measurement data from one or more measurement collector or agencies that are arranged in described physical space, described one or more measurement collectors or agency are same or different from the one or more assemblies in the described one or more assemblies that use in described cordless communication network;

In one or more performance index of described computing machine or the described cordless communication network of server prediction, wherein this prediction is based on from described one or more measurement collectors or agency's the described measurement data and the modeled attribute of at least one assembly in described one or more assembly and carries out; And

Change the setting or the configuration of at least one assembly of described cordless communication network based on the instruction of sending from described computing machine or server.

2. method according to claim 1, the specific expression in wherein said position is three-dimensional.

3. method according to claim 1, wherein said data acquisition are measured collector or the agency is of portable form or fixing.

4. method according to claim 1 further is included in the step of for good and all fixing described measurement collector or agency in the described physical space.

5. method according to claim 1, wherein the described performance index of predicting in described prediction steps are from by selecting the following group of forming: throughput, the bit error rate, packet awaits time, packet jitter, code element shake, service quality, security, overlay area, bandwidth, bit error rate, packet-error-rate, the frame bit error rate, packet loss, queueing delay, return interval, capacity, signal level, interference level, bandwidth-delay product, switching delay time, signal-to-noise ratio, signal noise ratio, physical equipment price and cost information.

6. method according to claim 1, wherein the measurement data that receives in described receiving step is manually to obtain.

7. method according to claim 1, wherein the described measurement data that receives in described receiving step is to obtain automatically.

8. method according to claim 1 further comprises the step of storing described measurement data.

9. method according to claim 1 further comprises the step of upgrading described computerized model.

10. method according to claim 9, wherein said step of updating may further comprise the steps:

Be identified for the assembly of described communication network from a plurality of different modeled assemblies, described modeled assembly comprises the description and the attribute of specific components; With

Determine the position of a plurality of different assemblies in described physical space in the described computerized model.

11. method according to claim 10, wherein said step of updating further comprises, on the determined position, determine the step of the direction of at least one assembly in described second determining step, this at least one assembly is determined in described first determining step.

12. method according to claim 1, the direction of the described assembly of identification on the described position of wherein said computerized model in described physical space, and described prediction steps is used described direction.

13. method according to claim 1, wherein said computerized model comprise the one or more targets that produce noise or interference, described noise or interference are the attributes as described one or more targets of the factor in the described prediction steps.

14. method according to claim 1, wherein described one or more performance index of predicting in described prediction steps are to predict on the forward direction of described cordless communication network.

15. method according to claim 1, wherein described one or more performance index of in described prediction steps, predicting be described cordless communication network oppositely on prediction.

16. method according to claim 1 further comprises the step of specified data host-host protocol, wherein said prediction steps uses the established data host-host protocol as the factor when the described one or more performance index of prediction.

17. method according to claim 1 further comprises the step of the offered load of determining described cordless communication network, wherein said prediction steps is used determined offered load when the described one or more performance index of prediction.

18. method according to claim 1 further comprises the storage or the step of visualization data, the comparison of this data representation measurement result and prediction.

19. method according to claim 1 further comprises the storage or the step of visualization data, the logic of this data representation networking component connect or the physical location of networking component in one or both.

20. the system or equipment analyzing and adjust cordless communication network comprises:

Computing machine or server, be used for producing or using the computerized model of the cordless communication network that is positioned at physical space, described computerized model provides the specific expression in the one or more position in place layout, building layout, regional feature or the radio frequency features, the position of one or more assemblies in described physical space of using in the described cordless communication network of described computerized Model Identification, described computerized model has the modeled attribute of at least one assembly in described one or more assembly;

Operation maybe can move in described physical space one or more measurement collectors or agency, it sends measurement data to described computing machine or server, and described computing machine or server are predicted one or more performance index of described cordless communication network based on the described modeled attribute of at least one assembly in described measurement data and the described one or more assembly; And described computing machine or server can send one or more assemblies that described cordless communication network is given in instruction, and described instruction is changed the setting of described at least one assembly or configuration.

21. system or equipment according to claim 20, wherein the specific expression in position is three-dimensional.

22. system or equipment according to claim 20, wherein said measurement collector or the agency is of portable form or fixing.

23. according to the system or equipment shown in the claim 20, wherein said measurement collector or agency are permanently fixed in the described physical space.

24. system or equipment according to claim 20, wherein the described performance index by described computing machine or server prediction are to select from the following group of forming: throughput, the bit error rate, the packet awaits time, packet jitter, the code element shake, service quality, security, the overlay area, bandwidth, bit error rate, packet-error-rate, the frame bit error rate, packet loss, queueing delay, return interval, capacity, signal level, interference level, bandwidth-delay product, the switching delay time, signal-to-noise ratio, signal noise ratio, physical equipment price and cost information.

25. system or equipment according to claim 20 further comprises the memory storage of storing described measurement data.

26. system or equipment according to claim 20 is wherein stored described computerized model at least one server, wherein said at least one server is same or different from described computing machine or server.

27. system or equipment according to claim 26 wherein store described computerized model in a plurality of servers, and described a plurality of server can be communicated by letter each other.

28. system or equipment according to claim 27, wherein said a plurality of servers have the relation of level each other.

29. system or equipment according to claim 26 further comprises at least one portable client device, described at least one portable client device can with described at least one server communication.

30. system or equipment according to claim 28, wherein said system or equipment comprise a plurality of portable client devices.

31. method according to claim 20 or equipment further comprise the storage medium or the display that are respectively applied for storage or visualization data, the comparison of this data representation measurement result and prediction.

32. method according to claim 20, further comprise be respectively applied for storage the logic of visualization networking component connects or the physical location of networking component in one or both storage medium or displays.

33. the method analyzing and adjust cordless communication network may further comprise the steps:

Produce or use the computerized model of the cordless communication network in the physical space with computing machine or server, described communication network is deployed in the described physical space, described computerized model provides the place layout, building model, the specific expression in one or more positions in topographic(al) feature or the radio frequency features, the position of one or more assemblies in described physical space of using in the described cordless communication network of described computerized Model Identification, described computerized model has the modeled attribute of at least one assembly in described one or more assembly;

The file of download or input measurement data is to described computing machine or server, and wherein said measurement data obtains from described physical space or from described cordless communication network;

One or more performance index of described cordless communication network are provided or are provided based on the described modeled attribute of at least one assembly in described measurement data and the described one or more assembly; And

Change the setting or the configuration of at least one assembly of described cordless communication network based on the instruction of sending from described computing machine or server.

34. method according to claim 33, wherein said measurement data are to obtain from portable or fixing measurement collector or agency.

35. method according to claim 33 further comprises the storage or the step of visualization data, the comparison of this data representation measurement result and prediction.

36. method according to claim 33 further comprises the storage or the step of visualization data, the logic of this data representation networking component connect or the physical location of networking component in one or both.

37. the position specifying method analyzing and adjust communication network may further comprise the steps:

Produce or use the computerized model that is positioned at the communication network in the physical space with computing machine or server, described computerized model provides the specific expression in the one or more position in place layout, building layout, topographic(al) feature or the radio frequency features, the position of one or more assemblies in described physical space of using in the described communication network of described computerized Model Identification, described computerized model has the modeled attribute of at least one assembly in described one or more assembly;

At described computing machine or server, receive measurement data from one or more measurement collectors that are arranged in described physical space or agency, described one or more measurement collectors or agency are same or different from one or more in the described one or more assemblies that use in described communication network;

Use one or more performance index of described computing machine or the described communication network of server prediction, wherein this prediction described modeled attribute of being based at least one assembly in described measurement data and the described one or more assembly carries out; And

Change the setting or the configuration of at least one assembly of described communication network based on the instruction of sending from described computing machine or server.

38. according to the described method of claim 37, the specific expression in wherein said position is three-dimensional.

39. according to the described method of claim 37, wherein said measurement collector or the agency is of portable form or fixing.

40., further be included in the step of for good and all fixing described measurement collector or agency in the described physical space according to the described method of claim 37.

41. according to the described method of claim 37, wherein described one or more performance index of predicting in described prediction steps are from by selecting the following group of forming: wireless signal strength density, connectivity, the network throughput, bit error rate, the frame bit error rate, signal-to-noise ratio, signal noise ratio, the frames per second resolution, portfolio, capacity, signal intensity, throughput, the bit error rate, the packet awaits time, packet jitter, the code element shake, service quality, security, area coverage, bandwidth, the server identification parameter, the sender identification parameter, the best server position, the sender location parameter, toll data, network performance parameter, C/I, C/N, the main body loss, floor height, apart from floor level, the noise figure, safety covers the position, the propagation loss factor, arrive angle, multipath component, the multipath parameter, antenna gain, the noise level reflectivity, surfaceness, path loss model, decay factor, the throughput performance index, packet-error-rate, return interval, packet loss, queueing delay, signal level, interference level, service quality, bandwidth-delay product, the switching delay time, loss of signal, data loss, the service-user number, customer density, suitably cover the position, switching position or zone, suitable throughput position, E _c/ Io, system performance parameter, equipment price, maintenance and cost information, class of subscriber or subclass, user type, position location, more than these all are absolute value or relative value.

42. according to the described method of claim 37, wherein the measurement data that receives in described receiving step is manually to obtain.

43. according to the described method of claim 37, wherein the described measurement data that receives in described receiving step is to obtain automatically.

44., further comprise the step of storing described measurement data according to the described method of claim 37.

45., further comprise the step of upgrading described computerized model according to the described method of claim 37.

46. according to the described method of claim 45, wherein said step of updating may further comprise the steps:

Be identified for the assembly of described communication network from a plurality of different modeled assemblies, described modeled assembly comprises the description and the attribute of specific components; With

Determine the position of a plurality of different assemblies in described space in the described computerized model.

47. according to the described method of claim 46, wherein said step of updating further comprises, on the determined position, determine the step of the direction of at least one assembly in the step of described definite position, this at least one assembly is determined in described definite assembly step.

48. according to the described method of claim 37, the direction of the one or more assemblies on the described position of wherein said computerized model in described physical space in the described one or more assemblies of identification, and described prediction steps is used described direction.

49. according to the described method of claim 37, wherein said computerized model comprises the one or more targets that produce noise or interference, described noise or interference are the attributes as described one or more targets of the factor in the described prediction steps.

50. according to the described method of claim 37, wherein described one or more performance index of predicting in described prediction steps are to predict on the forward direction of described communication network.

51. according to the described method of claim 37, wherein described one or more performance index of in described prediction steps, predicting be described communication network oppositely on prediction.

52. according to the described method of claim 37, further comprise the step of specified data host-host protocol, the factor when wherein said prediction steps is used the established data host-host protocol as the described performance index of prediction.

53. according to the described method of claim 37, further comprise the step of the offered load of determining described communication network, wherein said prediction steps is used the offered load of determining when the described one or more performance index of prediction.

54., further comprise the step of storage or visualization data, the comparison of this data representation measurement result and prediction according to the described method of claim 37.

55. according to the described method of claim 37, further comprise the storage or the step of visualization data, the logic of this data representation networking component connect or the physical location of networking component in one or both.

56. position particular system or the equipment analyzing and adjust communication network comprises:

Computing machine or server, be used for producing or using the computerized model of the communication network that is positioned at physical space, described computerized model provides the specific expression in the one or more position in place layout, building layout, regional feature or the radio frequency features, the position of one or more assemblies in described physical space of using in the described communication network of described computerized Model Identification, described computerized model has the modeled attribute of at least one assembly in described one or more assembly;

Be positioned at the one or more measurement collectors or the agency of described physical space, its acquisition also sends measurement data to described computing machine or server, and described computing machine or server are predicted one or more performance index of described communication network based on the modeled attribute of at least one assembly in described measurement data and the described one or more assembly; And described computing machine or server can send one or more assemblies that described communication network is given in instruction, and described instruction is changed the setting of at least one assembly or configuration.

57. according to the described system or equipment of claim 56, the specific expression in wherein said position is three-dimensional.

58. according to the described system or equipment of claim 56, wherein said measurement collector or the agency is of portable form or fixing.

59. according to the system or equipment shown in the claim 56, wherein said measurement collector or agency are permanently fixed to the position in the described physical space.

60. according to the described system or equipment of claim 56, wherein said one or more performance index are to select from the group of following composition: wireless signal strength density, connectivity, the network throughput, bit error rate, the frame bit error rate, signal-to-noise ratio, signal noise ratio, the frames per second resolution, portfolio, capacity, signal intensity, throughput, the bit error rate, the packet awaits time, packet jitter, the code element shake, service quality, security, area coverage, bandwidth, the server identification parameter, the sender identification parameter, the best server position, the sender location parameter, toll data, network performance parameter, C/I, C/N, the main body loss, floor height, apart from floor level, the noise figure, safety covers the position, the propagation loss factor, arrive angle, multipath component, the multipath parameter, antenna gain, the noise level reflectivity, surfaceness, path loss model, decay factor, the throughput performance index, packet-error-rate, return interval, packet loss, queueing delay, signal level, interference level, service quality, bandwidth-delay product, the switching delay time, loss of signal, data loss, the service-user number, customer density, suitably cover the position, switching position or zone, suitable throughput position, E _c/ Io, system performance parameter, equipment price, maintenance and cost information, class of subscriber or subclass, user type, position location, more than these all are absolute value or relative value.

61., further comprise the memory storage of storing described measurement data according to the described system or equipment of claim 56.

62., wherein in can being same or different from least one server of described computing machine or server, store described computerized model according to the described system or equipment of claim 56.

63. according to the described system or equipment of claim 62, wherein store described computerized model in a plurality of servers, wherein said a plurality of servers can be communicated by letter each other.

64. according to the described system or equipment of claim 63, wherein said a plurality of servers have the relation of level each other.

65. according to the described system or equipment of claim 62, further comprise can with at least one portable client device of described at least one server communication.

66. according to the described system or equipment of claim 64, wherein said system comprises a plurality of portable client devices.

67., further comprise the storage medium or the display that are respectively applied for storage or visualization data, the comparison of this data representation measurement result and prediction according to described method of claim 56 or equipment.

68. according to the described method of claim 56, further comprise be respectively applied for storage the logic of visualization networking component connects or the physical location of networking component in one or both storage medium or displays.

69. the method analyzing and adjust cordless communication network may further comprise the steps:

Produce the computerized model of the cordless communication network in the physical space with computing machine, described cordless communication network is deployed in the described physical space, described computerized model provides the specific expression in position and discerns the position of one or more assemblies in described physical space, these one or more assemblies participate in described cordless communication network and upload transmission information, and described computerized model has the modeled attribute of at least one assembly in described one or more assembly;

At described computing machine, receive measurement data from the one or more measurement collectors that are arranged in described physical space;

In one or more performance index of the described cordless communication network of described computer forecast, wherein this prediction is based on from the described measurement data of described one or more measurement collectors and the modeled attribute of at least one assembly in described one or more assembly and carries out; And

Change based on the instruction of sending and to participate in the setting that described cordless communication network is uploaded at least one assembly of transmission information from described computing machine.

70., wherein will instruct to send to participate in described at least one assembly that described cordless communication network is uploaded transmission information, to upgrade the relevant setting of at least one assembly from described computing machine based on prediction according to the described method of claim 69.

71. according to the described method of claim 69, wherein instruction is based on the performance index of one or more predictions of described cordless communication network.

72. according to the described method of claim 69, wherein change based on described instruction setting, participate in cordless communication network and upload at least one assembly of transmission information and comprise the network equipment, and wherein will instruct and be provided to the network equipment, thereby improve the performance of cordless communication network from computing machine based on the performance index of one or more predictions of described cordless communication network.

73., wherein use snmp protocol communication that described instruction is transferred to from described computing machine and participate in cordless communication network and upload described at least one assembly of transmission information according to the described method of claim 69.

74. according to the described method of claim 69, wherein the performance index based on one or more predictions of described cordless communication network provide instruction, thereby improve the performance of network.

75., wherein participate in cordless communication network and upload at least one assembly of transmission information and comprise the network equipment according to the described method of claim 69.

76. according to the described method of claim 75, wherein the network equipment comprises WAP.

77. a system that analyzes and adjust cordless communication network comprises:

Computing machine, described computing machine is carried out the computerized model that is positioned at the cordless communication network in the physical space, described computerized model provides the specific expression in position and discerns the position of one or more assemblies in described physical space, described one or more assembly participates in described cordless communication network and uploads transmission information, and described computerized model has the modeled attribute of at least one assembly in described one or more assembly;

The one or more measurement collectors that in described physical space, move, it sends to described computing machine with measurement data, the described modeled attribute of at least one assembly of described computer based in described measurement data and described one or more assembly is predicted one or more performance index of described cordless communication network, wherein said computing machine is constructed to instruction sent to and participates in one or more assemblies that described cordless communication network is uploaded transmission information, so that the setting of at least one assembly is changed.

78. according to the described system of claim 77, wherein will instruct to send to participate in described at least one assembly that described cordless communication network is uploaded transmission information, to upgrade the relevant setting of at least one assembly from described computing machine based on one or more performance index of described cordless communication network.

79. according to the described system or equipment of claim 77, wherein instruction is based on one or more performance index of described cordless communication network.

80. according to the described system or equipment of claim 77, wherein change based on described instruction setting, participate in cordless communication network and upload at least one assembly of transmission information and comprise the network equipment, and wherein will instruct and be provided to the network equipment, thereby the performance of improvement cordless communication network from computing machine based on one or more performance index of described cordless communication network.

81., wherein use snmp protocol communication that described instruction is uploaded described at least one assembly of transmission information from described computing machine or Server Transport to participating in cordless communication network according to the described system or equipment of claim 77.

82. according to the described system or equipment of claim 77, wherein the one or more performance index based on described cordless communication network provide instruction, thereby improve the performance of network.

83., wherein participate in cordless communication network and upload at least one assembly of transmission information and comprise the network equipment according to the described system or equipment of claim 77.

84. 3 described system or equipments according to Claim 8, the specific expression in wherein said position is three-dimensional.

Images