CN101119314A - A control system, device and method for multi-mode terminal service flow - Google Patents

Abstract

The invention provides a system, device and method for controlling service flow of a multi-mode terminal. The system includes a plurality of communication networks composed of a core network and an access network, and at least one multi-mode terminal, the multi-mode terminal is equipped with a plurality of wireless network interfaces matching the communication network, and the access network of the communication network Among them, a network context monitoring module is included; the multi-mode terminal includes a device context monitoring submodule, a service identification module, a decision engine module, and a switching execution module. It achieves minimum cost overhead and maximum resource utilization.

Description

A control system, device and method for multi-mode terminal service flow

technical field

The present invention relates to the field of packet network technology, in particular to a control system, device and method for multi-mode terminal service flow.

Background technique

Next Generation Network (NGN) is a network based on an all-IP architecture that integrates multiple access methods. With the progress of microelectronics technology, IPv6 technology and wireless communication processing technology, future terminal equipment will be a multi-mode communication terminal equipped with various wireless transceiver devices, which can simultaneously access multiple networks, thus bringing people For greater ease of use.

As shown in FIG. 1 , it is a diagram of an application scenario of a future network. Among them, the multimode terminal can be directly connected with the general packet radio service (General Packet Radio Service, GPRS) and the universal mobile telecommunications system (Universal Mobile Telecommunications System, UMTS) network, and can also be connected with the personal digital assistant (Personal Digital Assistant) through the infrared interface. , PDA) connected to each other, using the PDA to realize interconnection with the Internet. If three sessions are running in parallel on the multi-mode terminal, namely voice, e-mail and file download, since these three sessions have different service characteristics and service quality requirements, the transmission scheme is generally as follows. For the voice service, the service should be selected The UMTS network with the best quality can be used for transmission, e-mail can be transmitted using the GPRS network, and the file download business with a large amount of data can be accessed through the PDA in the same LAN with the lowest fee and the largest bandwidth Wireless LAN (Wireless LAN). , WLAN) network for transmission. In addition, because the user's movement may cause the WLAN network to be unavailable, the PDA should notify the multimode terminal at this time, so that the multimode terminal can switch the local file download session to the GPRS network to continue downloading in time.

That is to say, the future network will be a network integrating multiple wireless access modes, and the future terminal will be a terminal equipped with multiple wireless transceiver devices. Therefore, the popularization of multi-mode terminals is an inevitable trend, and multi-mode mobile communication will have important application value in military, commercial, civilian, medical and other scenarios.

However, in various application scenarios, when the multi-mode terminal moves or the network situation changes, it may cause the redistribution or switching of the service flow (also called service data flow) on the multi-mode terminal, and the traditional mobile IP cannot meet the requirements. The above requirements (see D.Johnson, C.Perkins, J.Arkko, "MobilitySupport in IPv6", IETF RFC3775), at this time, in a complex environment where multiple wireless networks converge, operators cannot provide dynamic support for multi-mode end users. When the network status and service flow status change, the service flow cannot be redirected to the new optimal network interface without restarting or interrupting the upper-layer business, and there is no guarantee Quality of service for users.

The Chinese invention patent with application number 200610088980.0 discloses a fuzzy logic-based vertical switching control system and control method, which includes: fuzzy module, fuzzy logic reasoning module, defuzzification module, knowledge base and configuration management module. The fuzzy decision-making algorithm comprehensively evaluates the network link quality, bandwidth, price, battery power, mobile speed and user preference of the mobile node (MN), and selects the optimal network for handover. The method provided by the invention solves the decision-making problem of optimal network selection in the process of vertical switching of mobile equipment with multiple network access technologies in a network environment with heterogeneous overlapping coverage.

However, although a multi-mode terminal is mentioned in this invention, a real multi-mode terminal is not realized. Only one wireless transceiver interface is available in the communication process, and the switching process is traditional vertical switching. This invention also does not provide how to realize the switching of a single service flow on multiple simultaneously available wireless interfaces of the multi-mode terminal. The user dynamically selects the optimal access network, and redirects the service flow to the new optimal network interface.

Contents of the invention

The purpose of the present invention is to provide a multi-mode terminal service flow control system, device and method for mobile multi-mode devices in future broadband wireless mobile communication networks and application scenarios where multiple devices cooperate with each other, so as to help terminals realize minimum Cost overhead and maximum resource utilization.

In order to achieve the above object, the present invention provides a multi-mode terminal service flow control system, including multiple communication networks composed of core networks and access networks, and at least one multi-mode terminal, the multi-mode terminal is equipped with multiple The wireless network interface matched with the communication network, the access network of the communication network includes a network context monitoring module; the multi-mode terminal includes a device context monitoring submodule, a service identification module, a decision engine module, and a switching execution module, wherein :

The network context monitoring module is used to monitor the network context parameters of the access network within the time period T, and pass the monitoring results to the decision engine module after being statistically averaged;

The device context monitoring module is used to monitor the device context parameters of the multi-mode mobile terminal within the time period T, and pass the monitoring results to the decision engine module after statistical average processing;

The service identification module is used to perceive service flow characteristic parameters, and feed back the obtained service flow characteristic parameters to the decision engine module;

The decision-making engine module is used for network context parameters, device context parameters and service flow characteristic parameters to analyze the service flow, establish a mapping relationship between the service flow and the interface, and select a wireless network interface for the service flow;

The multi-mode terminal also includes a switching execution module, which is used to start a switching signaling process according to the wireless interface selected by the decision engine module for a single service flow, and allocate or switch the service flow to the wireless transceiver interface.

The multi-mode terminal further includes an execution result feedback module, which is used to feed back the execution status of previous decision results and sensitivity to parameters to the decision engine module, and the decision engine module revises the next decision based on the feedback.

The business identification module is also used to statistically analyze and record historical characteristics of business flows using machine learning methods, and to establish a classification mechanism for different types of business flows to improve the efficiency of machine learning.

The network context parameters include available bandwidth parameters of the access network, network delay parameters and link signal quality parameters.

The network context monitoring module monitors available network bandwidth and network delay parameters through active measurement and improved packet pair algorithm.

The device context parameters include wireless interface utilization parameters, battery power parameters, and wireless network interface transceiver power parameters.

The characteristic parameters of the service flow include the characteristics of the service flow, the type of the service flow, and the degree of demand of the service flow on the quality of service;

The historical characteristics of the service flow are the historical characteristics of the service flow of data packet size and average rate parameters.

The decision engine module establishes the mapping relationship between the service flow and the interface by using a fuzzy multi-attribute decision-making method.

The decision engine module works in the mode of input, output and feedback;

Among them, the input quantity is the network context parameter, the device context parameter, and the service flow characteristic parameter;

The output volume is the mapping relationship between the service flow and the wireless network interface;

The feedback amount is obtained from the execution result feedback module, and is used to assist in correcting the next decision.

The decision-making basis of the decision-making engine module is a network cost function, and the network cost function is established by parameters such as network context parameters, device context parameters and service flow characteristics, and is expressed as follows:

fun _Net ＝W _net para _net +W _equip para _equip +W _flow para _flow

Among them, para _net represents the network context parameter, W _net represents the weight of the network context parameter, para _equip represents the device context parameter, W _equip represents the weight of the device context parameter, para _flow represents the service flow characteristic parameter, W _flow represents the service flow characteristic parameter Weights;

The sum of the weights is 1, that is, W _net +W _equip +W _flow =1, and the size of the weight is obtained according to previous experience information, and fine-tuned according to the parameters fed back by the execution result feedback module.

The multi-mode mobile terminal is equipped with multiple care-of-addresses, that is, a main care-of-address and multiple sub-care-of-addresses. The multiple sub-care-of-addresses have a corresponding relationship with the main care-of-address. According to the result of the binding update message comparison, the matching service Streams are directed to the new wireless network interface.

The plurality of communication networks are GPRS communication networks, UMTS communication networks, infrared networks, WLAN networks, WiMAX communication networks, Bluetooth communication networks, or a combination of two or more;

The multi-mode terminal includes a plurality of corresponding wireless network interfaces, which are respectively a combination of two or more of GPRS interface, UMTS interface and infrared interface, WLAN network interface, WiMAX communication network interface, and Bluetooth communication network interface;

The multi-mode terminal includes a combination of two or more of email service flows, file download service flows, voice service flows, and multimedia service flows.

In order to achieve the purpose of the present invention, a communication network for multi-mode terminal service flow control is also provided, which is composed of a core network and an access network. The access network includes a network context monitoring module, and the network context monitoring module is used for The network context parameters of the access network are monitored within the time period T, and the monitoring results are statistically and averaged before being delivered to the multi-mode terminal.

The communication network is a combination of two or more of GPRS communication network, UMTS communication network, infrared network, WLAN network, WiMAX communication network, and Bluetooth communication network.

In order to achieve the purpose of the present invention, a multi-mode terminal for service flow control is further provided, including a device context monitoring submodule, a service identification module, a decision engine module, and a switching execution module, wherein:

The device context monitoring module is used to monitor the device context parameters of the multi-mode mobile terminal within the time period T, and pass the monitoring results to the decision engine module after statistical average processing;

The service identification module is used to perceive service flow characteristic parameters, and feed back the obtained service flow characteristic parameters to the decision engine module;

The decision-making engine module is used for network context parameters, device context parameters and service flow characteristic parameters to analyze the service flow, establish a mapping relationship between the service flow and the interface, and select a wireless network interface for the service flow;

The multi-mode terminal also includes a switching execution module, which is used to start a switching signaling process according to the wireless interface selected by the decision engine module for a single service flow, and allocate or switch the service flow to the wireless transceiver interface.

The multi-mode terminal for business flow control further includes an execution result feedback module, which is used to feed back the execution status of previous decision results and the sensitivity to parameters to the decision engine module, and the decision engine module revises the result according to the feedback. One decision.

The business identification module is also used to statistically analyze and record historical characteristics of business flows using machine learning methods, and to establish a classification mechanism for different types of business flows to improve the efficiency of machine learning.

The multi-mode terminal includes a plurality of corresponding wireless network interfaces, which are respectively a combination of two or more of GPRS interface, UMTS interface and infrared interface, WLAN network interface, WiMAX communication network interface, and Bluetooth communication network interface;

The multi-mode terminal includes a combination of two or more of email service flows, file download service flows, voice service flows, and multimedia service flows.

In order to realize the purpose of the present invention, a method for controlling service flow of a multi-mode terminal is further provided, comprising the following steps:

Step A, within the time period T, the access network monitors its network context parameters and transmits them to the multi-mode terminal; the multi-mode terminal monitors its device context parameters, and then performs statistical average processing on the monitored network context parameters and device context parameters; at the same time , the characteristic parameter of the service flow perceived by the multi-mode terminal, the obtained characteristic parameter of the service flow;

Step B, the multi-mode terminal analyzes the service flow according to the network context parameter, the device context parameter and the service flow characteristic multiple parameters, establishes a mapping relationship between the service flow and the interface, and selects a wireless network interface for the service flow;

In step C, a switching signaling process is started according to the wireless interface selected by the multi-mode terminal for a single service flow, and a service flow is allocated or switched to a wireless transceiver interface.

The method for controlling the service flow of the multi-mode terminal also includes the following steps:

In step D, the multi-mode terminal revises the next decision according to the implementation of previous decision results and sensitivity to parameters.

Described step A comprises the following steps:

Step A1, within the time period T, the access network of the communication network monitors multiple network context parameters such as available bandwidth, network delay and link signal quality, and transmits the monitoring results to the multi-mode terminal after statistical averaging processing decision engine module;

Step A2, within the time period T, the device context detection module of the multi-mode terminal monitors multiple device context parameters of its wireless interface utilization, battery power, and wireless network interface transmission and reception power, and transmits the monitoring results after statistical average processing Decision engine module for multimode terminals;

Step A3, the service identification module of the multi-mode terminal perceives multiple service flow characteristic parameters such as the characteristics of the new service flow, the type of the service flow, and the degree of demand for service quality of the service flow, and uses machine learning methods to statistically analyze and record the historical characteristics of the service flow, And for different types of business flows, a classification mechanism is established, and the obtained characteristic parameters of the business flows are fed back to the decision engine module of the multi-mode terminal.

In step A1, the network context monitoring module monitors the network monitoring available bandwidth and network delay parameters through active measurement and improved packet algorithm monitoring, including the following steps:

Step A11, constructing a detection data packet on the multi-mode terminal at the sending end, the detection data packet is divided into a large packet and a small packet;

Step A12, the multi-mode terminal at the sending end sends multiple pairs of detection data packets, and the order of sending the detection packet pairs is that the large packet pairs and the small packet pairs alternate with each other, and these detection data packets are stamped with time stamps and sending numbers when sent by the sending end;

Step A13, the multimode terminal at the receiving end receives the detection data packet, and stamps a time stamp on each detection data packet received;

Step A14, the multi-mode terminal at the receiving end checks the number of the received detection data packet, and judges whether the sequence of the received detection data packet is consistent with the order sent by the multi-mode terminal at the sending end; if not, discards the detection data packet and goes to step A11, start again; if there is no disorder, then go to step A15;

Step A15, according to the time stamps of the received detection data packets, respectively calculate the time interval between the packet pairs, that is, the time difference between each pair of detection data packets arriving at the receiving end;

Step A16, judging whether the time interval of checking the small packet pair is greater than the time interval of the large packet pair; if the time interval between the small packet pairs is greater than the time interval between the large packet pairs, then this measurement is invalid, return to step A11, and start again; Otherwise, go to step A17;

Step A17, check whether the time interval of the large packet pair is the same, if the time interval of the large packet pair is the same, then this measurement is invalid, return to step A11, and start again; otherwise, enter step A18 to calculate the effective bandwidth of the network, and step A110 to calculate the network time extend;

Step A18, compare the size of the time interval of the large packet pair, and take out the smaller time interval;

Step A19, calculating the network effective bandwidth B=S/T _min , wherein S is the data packet size of the large detection packet, and T _min is the smaller time interval of the large packet pair;

Step A110, compare the transmission time interval of the sent data packets, that is, the difference between the receiving timestamp and the sending timestamp of each data packet, find out the data packet with the smallest transmission time interval, and take out the sending timestamp of this data packet and receive timestamp;

Step A111, according to the retrieved sending timestamp and receiving timestamp, use the linear programming method to reduce the calculated network delay error caused by the clock asynchrony of the sending host and the receiving host, and calculate a relatively accurate network delay.

The following steps are also included after step A111:

Step A112, on the access network, collect the link signal quality in real time by calling the function interface provided by the system;

Step A113 , within the time period, process the calculated monitoring results of network effective bandwidth, network delay, and link signal quality parameters by statistical averaging method, and transmit them to the multi-mode terminal.

In the step A2, the device context monitoring module monitors multiple device context parameters of the multi-mode mobile terminal, and processes the monitoring results through statistical averaging, including the following steps:

Step A21, on the mobile terminal, call the function interface provided by the system through the device context monitoring module, and collect the device interface utilization rate, battery power and wireless network interface transceiver power parameters of the mobile terminal in real time;

Step A22, within the time period T, process the collected parameters such as device interface utilization, battery power, and wireless network interface transceiver power through a statistical averaging method.

In step B, the statistical analysis records the historical characteristics of the business flow, specifically including the following steps

The business identification module adopts the method of machine learning to analyze and record the type of business flow, the characteristics of the business flow and the demand of the service quality parameters of the business flow, and statistically analyze the historical characteristics of the business flow of the recorded data packet size and average rate parameters.

The large packet is an IP header plus an ICMP timestamp request and response message, and a data packet with additional data;

The small packet is a request and response message with an IP header plus an ICMP time stamp, without any additional data.

In the step C, the establishment of the mapping relationship between the service flow and the interface includes the following steps:

Use the fuzzy multi-attribute decision-making method to establish a reasonable mapping relationship between business flows and interfaces, among which,

The basis of the decision-making method is a network cost function, and the network cost function is established by parameters such as network context parameters, device context parameters and service flow characteristics, and can be expressed as follows:

fun _Net ＝W _net para _net +W _equip para _equip +W _flow para _flow

Among them, para _net represents the network context parameter, W _net represents the weight of the network context parameter, para _equip represents the device context parameter, W _equip represents the weight of the device context parameter, para _flow represents the service flow characteristic parameter, W _flow represents the service flow characteristic parameter Weights;

The sum of the weights is 1, that is, W _net +W _equip +W _flow =1, and the size of the weight is obtained according to previous experience information, and fine-tuned according to the parameters fed back by the execution result feedback module.

In the step D, the switching includes the following steps:

The multi-mode mobile terminal is equipped with multiple care-of-addresses, that is, a main care-of-address and multiple sub-care-of-addresses. The multiple sub-care-of-addresses have a corresponding relationship with the main care-of-address. According to the result of the binding update message comparison, the matching service Streams are directed to the new wireless network interface.

The beneficial effect of the present invention is that: the multi-mode terminal service flow control system, device and method of the present invention can be based on various service flow characteristics, service flow requirements for service quality parameters, current loads of different access networks, and device interfaces. Factors such as utilization rate and battery power, reasonably allocate service flows among multiple network interfaces of the same mobile terminal to obtain the minimum cost overhead and maximum resource utilization. When characteristics change, it can sense these changes in time and realize fast switching of a single service flow between different interfaces.

Description of drawings

Figure 1 is an application scenario diagram of a typical multi-mode terminal service system;

Fig. 2 is the schematic diagram of the control system of multi-mode terminal service flow of the present invention;

Fig. 3 is a flow chart of a method for controlling a multi-mode terminal service flow in the present invention;

Fig. 4 is a flow chart of the parameter monitoring process of the present invention;

Fig. 5 is a flow chart of the monitoring execution process of available bandwidth and network delay in the present invention.

Detailed ways

In order to make the purpose, technical solution and advantages of the present invention clearer, a system, device and method for controlling multi-mode terminal service flow of the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

The next-generation network based on the IPv6 protocol is a network based on an all-IP system. The construction of the entire network is composed of a core network and an access network. The core network is forwarded by routers. The entity names used in the entire network can be flexibly changed. For example, in a wireless local area network (Wireless Local Area Network, WLAN), a mobile user is called a mobile terminal or a mobile node, and a mobile terminal accesses a router through an access point (AccessPoint, AP); while in a CDMA network, a mobile terminal is a Access to the network through the base station.

In the present invention, a multi-mode terminal service flow control system, device and method are provided, which realize the service flow (service data flow) on the same multi-mode terminal by extending the mobile IP protocol (IPv6 protocol). ) separation, seamless switching of a single service flow from one wireless interface to another wireless interface, data forwarding and flow switching between multiple devices, and can also be implemented on multi-mode terminals according to service flow characteristics, service flow The demand for quality of service parameters, access network conditions, equipment utilization and other factors, and through the mutual cooperation relationship between multi-mode terminals, select the optimal wireless transceiver interface for business flows to achieve the minimum cost overhead and maximum At the same time, it can ensure that when the network status and service flow characteristics change, it can sense these changes in time, and realize the rapid switching of a single service flow between different interfaces.

Among them, the mobile IP protocol (IPv6 protocol, RFC3775 standard) is a three-layer international protocol for the next generation network. It marks the identity of the mobile node (Mobile Node, MN) through the home address (Home of Address, HoA) of Adrress, CoA) indicates the location of the mobile node. When the mobile node is in the home network, it communicates with the corresponding node (Corresponding Node, CN) through the home address. Mobile IPv6 protocol solves the roaming and switching problems of mobile nodes in the network.

The multi-mode terminal service flow control system in the present invention includes multiple communication networks composed of core networks and access networks, and multiple multi-mode terminals 1 .

As shown in Figure 1, the communication network of the present invention can be General Packet Radio Service (General Packer Radio Service, GPRS) network 2, or Universal Mobile Telecommunications System (Universal MobileTelecommunication System, UMTS) network 3, or infrared network 4, or It is a combination of two or more such as a Worldwide Interoperability for Microwave Access (WiMAX) (not shown), or a wireless local area network (Wireless Local Area Network, WLAN) (not shown) .

As a possible implementation mode, the embodiment of the present invention uses a system composed of GPRS network 2 , UMTS network 3 and infrared network 4 to illustrate the multi-mode terminal service flow control system of the present invention. However, it should be noted that the system described in the present invention is not limited to the above-mentioned communication network, and it may also be composed of existing or future communication networks that may be combined into a multi-mode communication network.

The multi-mode terminal 1 includes a sending-end multi-mode terminal and a receiving-end multi-mode terminal, and the multi-mode terminal 1 is equipped with a plurality of wireless network interfaces matching the communication network.

As a possible implementation manner, as shown in FIG. 2 , the multimode terminal 1 in the embodiment of the present invention is equipped with three wireless network interfaces, which are GPRS interface 27 , UMTS interface 28 and infrared interface 29 . Four types of service flows are concurrently running on the multimode terminal 1 , namely email (Email) service flow 210 , file download (FTP) service flow 220 , voice (VoIP) service flow 230 , and multimedia service flow 240 .

It should be understood that GPRS interface 27, UMTS interface 28 and infrared interface 29 and electronic mail (Email) service flow 210, file download (FTP) service flow 220, speech (VoIP) service flow 230 and multimedia service flow among the present embodiment 240 is used for example only, not as a limitation to the present invention. Corresponding to a WiMAX network (not shown) and a WLAN network (not shown), the wireless network interface may also be a WiMAX interface (not shown), a WLAN interface (not shown), etc., and the service flow may also be are other corresponding types of business flows.

As shown in Figure 2, the control system of the multimode terminal service flow of the present invention includes a network context monitoring module 21 in the access network; the multimode terminal 1 includes a device context monitoring module 22, a service identification module 23, a decision-making An engine module 24 , a switching execution module 25 , and an execution result feedback module 26 .

The network context monitoring module 21 is used to monitor multiple network context parameters such as the available bandwidth of the access network, network delay, and link signal quality within the time period T, and transmit the monitoring results after statistical average processing To the decision engine module 24 of the multimode terminal 1.

Wherein, preferably, the network context parameters such as available bandwidth and network delay are obtained through active measurement and monitoring using an improved packet pairing algorithm.

The device context monitoring module 22 is used to monitor multiple device context parameters such as the wireless interface utilization rate, battery power and wireless network interface transceiver power of the multi-mode mobile terminal within the time period T, and process the monitoring results through statistical averaging After that, it is passed to the decision engine module 24.

The service identification module 23 is used to perceive multiple service flow characteristic parameters such as the characteristics of the new service flow, the type of the service flow, and the demand degree of the service flow for the service quality, and adopts the machine learning method to statistically analyze and record the historical characteristics of the service flow, and For different types of business flows, a classification mechanism is established to improve the efficiency of machine learning, and the obtained characteristic parameters of the business flows are fed back to the decision engine module 24 .

The historical characteristics of the service flow are historical characteristics of the service flow with multiple parameters such as data packet size and average rate.

The decision-making engine module 24 is configured to conduct a business flow according to multiple parameters such as network context parameters, device context parameters, and service flow characteristics fed back by the network context monitoring module 21, the device context monitoring module 22, and the service flow identification module 23. Analysis, use the fuzzy multi-attribute decision-making method to establish the mapping relationship between the service flow and the interface, and select the wireless network interface for the service flow.

The decision engine module 24 works in the mode of input, output and feedback. Among them, the input quantity is the network context parameter, the device context parameter, and the service flow characteristic parameter; the output quantity is the mapping relationship between the service flow and the wireless network interface; and the feedback quantity is obtained from the execution result feedback module 26, which is used for auxiliary correction One decision.

The handover execution module 25 is configured to start the handover signaling process according to the wireless interface selected by the decision engine module 24 for a single service flow, and allocate or switch the service flow to the wireless transceiver interface.

The execution result feedback module 26 is used to feed back the execution status of previous decision results and the sensitivity to parameters to the decision engine module 24, and the decision engine module 24 revises the next decision according to the feedback.

In the multi-mode terminal service flow control system of the present invention, the network context monitoring module 21 and the device context monitoring module 22 pass the network context and device context parameters monitored within a certain period of time to the decision engine module 24 after statistical averaging processing. The business identification module 23 uses machine learning methods to learn these several types of business flows, perceive the characteristics of the business flows, the types of the business flows, and the degree of service quality requirements of the business flows, analyze and record the historical characteristics of the business flows, and analyze the business flows. Flow establishes a reasonable classification mechanism, and then transmits the analysis results of the business flow to the decision engine module 24. At the same time, the execution result feedback module 26 also feeds back the previous decision result to the decision engine module 24. The decision engine module 24 comprehensively considers Various parameters from the monitoring module and service identification module 23, and referring to the information delivered by the execution result feedback module 26, use the method of fuzzy multi-attribute decision-making to select the most reasonable wireless network interface for each service flow.

The working process of the multi-mode terminal service flow control system of the present invention is described in detail below, that is, the present invention provides a method for controlling multi-module terminal service flow, as shown in Figure 3, the method includes the following steps:

Step S100, within the time period T, the network context monitoring module 21 of the access network of the communication network monitors multiple network context parameters such as its available bandwidth, network delay and link signal quality, and transmits them to the multimode terminal 1; Terminal 1 monitors multiple device context parameters such as the utilization rate of its wireless interface, battery power, and wireless network interface transceiver power, and then performs statistical average processing on the monitored network context parameters and device context parameters; at the same time, multi-mode terminal 1 perceives new service flows characteristics of business flows, types of business flows, and service quality requirements of business flows, etc., use machine learning methods to statistically analyze and record historical characteristics of business flows, and establish a classification mechanism for different types of business flows to obtain The characteristic parameters of the service flow are fed back to the decision engine module 24 of the multimode terminal 1.

The network context monitoring process of the present invention is shown in Figure 4, and the specific steps are as follows:

Step S110, within the time period T, the network context monitoring module 21 of the access network of the communication network monitors multiple network context parameters such as its available bandwidth, network delay, and link signal quality, and performs statistical averaging on the monitoring results After that, it is passed to the decision engine module 24 of the multimode terminal 1.

As an implementable manner, multiple network context parameters such as the available bandwidth and network delay of the monitoring network of the communication network are obtained through active monitoring.

Among them, the active monitoring method is one of the two common IP network performance monitoring methods. It uses monitoring tools to actively generate monitoring traffic at selected monitoring points, injects it into the network, and monitors traffic according to the transmission of monitoring traffic. situation to analyze the performance of the network.

As an implementable manner, in the embodiment of the present invention, the available bandwidth monitoring of the access network of the communication network may be performed by the Cprobe method, the Pathload method, and the TOPP algorithm.

Preferably, in the present invention, the available bandwidth and network delay are obtained through active monitoring and improved packet pair algorithm monitoring.

The improved packet pair algorithm is to send a detection packet between the multimode terminal 1 and a network access point (Access Point, AP), and after the construction of the detection packet is completed on the multimode terminal 1, it is sent to the access point, To monitor the available bandwidth and network delay of the last hop.

This algorithm overcomes some disadvantages of the classic packet pair algorithm. The algorithm has a small amount of calculation and low complexity, and the accuracy of the obtained network bandwidth can meet the requirements of the system.

As shown in Figure 5, the steps of monitoring the available bandwidth and network delay in the embodiment of the present invention are as follows:

Step S111, constructing a detection data packet on the multi-mode terminal at the sending end, the detection data packet is divided into a large packet and a small packet;

Among them, a large packet refers to a data packet with an Internet Control Message Protocol (Internet Control Message Protocol, ICMP) timestamp request and response message added to the IP header, and a certain bit of data is attached.

In the embodiment of the present invention, the data attached to the large packet can be unfixed, and it can be adjusted appropriately according to the actual situation. The data is common data, and the network control information protocol is only added to the head of the data;

A small packet refers to a data packet with an ICMP timestamp request and response message added to the IP header, but without any additional data. Preferably, the size of the packet is 40 bytes.

Step S112, the multi-mode terminal at the sending end sends multiple pairs of detection data packets, and the order of sending the detection packet pairs is that the large packet pairs and the small packet pairs alternate with each other, and these detection data packets are stamped with time stamps and sending numbers when sent by the sending end;

Step S113, the multimode terminal at the receiving end receives the detection data packet, and stamps a time stamp on each detection data packet received;

Step S114, the multimode terminal at the receiving end checks the number of the received detection data packet, and judges whether the order of the received detection data packet is consistent with the order sent by the multimode terminal at the sending end; If the second measurement is invalid, discard the detection data packet, go to step S111, and start again; if there is no out-of-sequence phenomenon, then go to step S115;

Step S115, the network context monitoring module of the access network calculates the time interval between the packet pairs according to the time stamps of the received detection data packets, that is, the time difference between each pair of detection data packets arriving at the receiving end;

Step S116, the network context monitoring module of the access network judges whether the time interval for checking small packet pairs is greater than the time interval for large packet pairs; if the time interval between small packet pairs is greater than the time interval between large packet pairs, then this measurement is invalid , return to step S111, and start again; otherwise, enter step S117;

Since the data packets go through the same path, the transmission time of large packets should be longer than that of small packets, so the time interval between pairs of small packets should be smaller than the time interval between pairs of large packets. Therefore, if the time interval between pairs of small packets If it is greater than the time interval between large packet pairs, it means that there is interference, and this measurement is invalid.

Step S117, the network context monitoring module of the access network checks whether the time interval of the large packet pair is the same, if the time interval of the large packet pair is the same, then this measurement is invalid, return to step S111, and start again; otherwise, enter step S118 to calculate the network is valid Bandwidth B, and step S1110 calculates network delay D;

Since it is only an ideal state without any interference on the channel during data packet transmission, and this ideal state does not exist, the time spent in transmitting large packets must be different, that is, the large packet has a negative effect on it. Therefore, the network context monitoring module of the access network checks whether the time intervals of large packet pairs are the same. If the time intervals of large packet pairs are the same, there must be a problem, and this measurement is invalid.

Step S118, the network context monitoring module of the access network compares the size of the time interval of the large packet pair, and takes out a smaller time interval T _min ;

Step S119, the network context monitoring module of the access network calculates the effective bandwidth of the network, B=S/T _min , wherein S is the data packet size of the large detection packet, and T _min is the smaller time interval of the large packet pair;

Step S1110, the network context monitoring module of the access network compares the transmission time interval of the data packets sent, that is, the difference between the receiving time stamp and the sending time stamp of each data packet, finds out the data packet with the smallest transmission time interval, and Take out the sending timestamp T1 and receiving timestamp T2 of this packet;

Step S1111, the network context monitoring module of the access network uses the linear programming method to reduce the calculation network delay error caused by the asynchronous clocks of the sending host and the receiving host according to the sending timestamp T1 and receiving timestamp T2 taken out in step S1110 , calculate the relatively accurate network delay D;

There are two main sources of delay error: one is the error caused by monitoring the resolution, offset, and jitter of the host system clock, which affects the timing accuracy of the measurement, which is called clock error here; The monitoring error generated at the moment of sending and receiving packets is not real, which is called position error here.

In the embodiment of the present invention, the forward time delay and the reverse time delay are calculated by sending active detection data packets, and then the error is reduced by using a linear programming method.

Wherein, the linear programming method is a prior art, which is a method for eliminating errors. Similar methods for eliminating errors include graph programming methods, linear regression methods, etc., and linear programming methods are used in specific embodiments of the present invention to eliminate errors. Errors are for example only, not limitation of the present invention.

The link signal quality parameters are collected in real time by calling the function interface provided by the system.

Among them, the link signal quality parameters include parameters such as signal strength and link signal-to-noise ratio. These parameters can call the system function interface. For example, in the WLAN protocol, all frames received by the second layer correspond to a state information states structure. This structure includes parameters such as received signal strength and signal-to-noise ratio, so the current signal strength and signal-to-noise ratio can be obtained from the second-layer network card driver.

In addition, within a certain period of time T, the network context monitoring module 21 in the access network of the communication network performs statistical average processing on the calculated available network bandwidth, network delay, and collected link signal quality, and then performs these The parameters are passed to the decision engine module 24 of the multimode terminal 1 .

Step S120, within the time period T, the device context detection module of the multi-mode terminal 1 monitors multiple device context parameters such as its wireless interface utilization rate, battery power, and wireless network interface transceiver power, and performs statistical averaging on the monitoring results , and pass it to the decision engine module 24 of the multimode terminal 1.

By calling the system function interface, the device interface utilization, battery power, and wireless network interface transceiver power are collected in real time, and then, within the time period T, the obtained device interface utilization, battery power, and wireless network interface transceiver power are respectively For statistical averaging.

Among them, the acquisition of these parameters corresponds to different system function interfaces, such as the acquisition of device interface utilization. For WLAN networks, the hook function mechanism provided by Netfilter can be used to count the wireless device interface utilization.

Step S130, the service identification module 23 of the multi-mode terminal 1 perceives multiple service flow characteristic parameters such as the characteristics of the new service flow, the type of the service flow, and the degree of demand for service quality of the service flow, and uses machine learning methods to statistically analyze and record the history of the service flow characteristics, and establish a classification mechanism for different types of service flows, and feed back the obtained service flow characteristic parameters to the decision engine module 24 of the multimode terminal 1 .

When the multi-mode terminal 1 is connected to the access network or after the multi-mode terminal is connected to the access network, it perceives multiple service flow characteristic parameters such as the characteristics of the new service flow, the type of the service flow, and the degree of service quality demanded by the service flow. , use machine learning methods to statistically analyze and record the historical characteristics of business flows, and establish a classification mechanism for different types of business flows, improve the efficiency of machine learning, and feed back the obtained business flow characteristic parameters to the decision engine module 24 of the multi-mode terminal 1 .

The business identification module 23 of this multi-mode terminal 1 adopts machine learning method, for example, Bayesian classifier identifies and classifies business flow, perceives the feature of new business flow, wherein, Bayesian classifier passes training set (classified example Set) training (learning) to induce a classifier, and use the classifier to classify the unclassified data.

The historical characteristics of the service flow are historical characteristics of the service flow with multiple parameters such as data packet size and average rate.

In step S200, the decision engine module 24 of the multi-mode terminal 1 analyzes the service flow according to the feedback network context parameters, device context parameters, and service flow characteristics, and uses the fuzzy multi-attribute decision-making method to establish the relationship between the service flow and the interface. Reasonable mapping relationship among them, and select the optimal wireless network interface for business flow.

Among them, the fuzzy multi-attribute decision-making method is to use the existing decision-making information to sort a group of selected options in a certain way.

Including the following steps:

Step S210, acquiring decision information. Decision information includes two aspects: attribute weight and attribute value. In the embodiment of the present invention, the attribute values used by the decision engine module 24 are obtained from the network context monitoring module 21 , the device context monitoring module 22 and the service identification module 23 . The attribute weights are obtained from previous decision results, that is, from empirical information.

Step S220, integrating the decision information and sorting various proposals.

The decision engine module 24 of the multimode terminal 1 works in the mode of input, output and feedback. Among them, the input quantity is the network context parameter, the device context parameter, and the service flow characteristic parameter; the output quantity is the mapping relationship between the service flow and the wireless network interface; and the feedback quantity is obtained from the execution result feedback module 26, which is used for auxiliary correction One decision.

Preferably, as an implementable manner, in the embodiment of the present invention, the decision engine module 24 of the multimode terminal 1 makes a decision based on a network cost function. The network cost function is established by parameters such as network context parameters, device context parameters, and service flow characteristics, and can be expressed as follows:

fun _Net ＝W _net para _net +W _equip para _equip +W _flow para _flow (1)

Among them, para _net represents the network context parameter, W _net represents the weight of the network context parameter, para _equip represents the device context parameter, W _equip represents the weight of the device context parameter, para _flow represents the service flow characteristic parameter, W _flow represents the service flow characteristic parameter Weights.

The sum of the weights is 1, that is, W _net +W _equip +W _flow =1, and the size of the weight is obtained according to previous experience information, and fine-tuned according to the parameters fed back by the execution result feedback module 26 .

The decision engine module 24 of the multimode terminal 1 compares the calculated network cost function values, and uses the wireless transceiver interface corresponding to the minimum network cost function value to transmit the service flow, and then the decision engine module 24 of the multimode terminal 1 takes the decision The result is fed back to the switching execution module 25 of the multimode terminal 1 .

In step S300, a switching signaling process is started according to the wireless interface selected by the decision engine module 24 of the multi-mode terminal 1 for a single service flow, and a service flow is allocated or switched to a wireless transceiver interface.

The mobile IP solution in the prior art directs all service flows to one wireless access technology, which does not meet the requirement of dynamically allocating service flows with different characteristics to different wireless links for transmission.

Therefore, preferably, the present invention modifies the existing mobile IP solution, and subdivides the switching granularity from a single device to a single service flow, so as to realize efficient and rapid service flow redirection.

The multimode mobile terminal of the present invention is equipped with multiple care-of addresses, namely a main care-of address (mainCoA, mCoA) and multiple sub-care-of addresses (subsidiary CoA, sCoA), and the multiple sub-care-of addresses have a corresponding relationship with the main care-of address. During the entire service flow control process, the primary care-of address of the multimode terminal 1 remains unchanged, and the core network in each communication network can store information of all flows passing through itself, namely source address, destination address and flow label.

In the present invention, an address object is also added in the traditional binding update (Binding Update, BU) message, and this object includes the old child care-of address of the multimode terminal 1, the main care-of address of the multimode terminal 1, and the 1 The address of the communication node (Correspondence Node, CN).

When the binding update message arrives at the core network of the communication network, the core network of the communication network compares the stored service flow information with the address object in the binding update message, and then directs the matching service flow to the new wireless network interface superior.

Step S400, the execution result feedback module 26 of the multi-mode terminal 1 feeds back the execution status of previous decision results and the sensitivity to parameters to the decision engine module 24, and the decision engine module 24 revises the next decision based on the feedback.

The execution result feedback module 26 feeds back to the decision engine module 24 the execution status of the decision result based on the decision result of the decision engine module 24 one or more times before, and the mapping relationship between the business flow and the interface. The sensitivity of the flow to parameters such as network bandwidth and network delay is fed back to the decision engine module 24, so that the decision engine module 24 can refer to these feedback information when making decisions, fine-tune the weight of each parameter, and select the best solution for the business flow. Better wireless network interface.

The multi-mode terminal service flow control method of the present invention is based on various service flow characteristics, service flow types, service flow requirements for service quality parameters, network available bandwidth, network delay, device interface utilization, and battery power. Parameters, reasonably allocate service flows among multiple network interfaces of the same mobile terminal to achieve the minimum cost overhead and maximum resource utilization, and at the same time ensure that when the network conditions and service flow characteristics change, they can be sensed in time change, and realize fast switching of a single service flow between different interfaces.

Experiment shows that the result of decision-making is: email (Email) service flow 210 is distributed on the wireless network GPRS interface 27, voice (VoIP) service flow 230 and multimedia service flow 240 are distributed on the wireless network UMTS interface 28, file download ( FTP) service flow 220 is distributed to the wireless network infrared interface 29. Then, the handover execution module 25 starts the relevant signaling process according to the wireless transceiver interface selected by the decision engine module 24 for each service flow, executes the handover operation, allocates or switches a service flow to a wireless transceiver interface, and completes the service flow The allocation and switching process is completed without user participation, so that users can enjoy the minimum cost overhead and maximum resource utilization.

Of course, the present invention can also have other various embodiments, and those skilled in the art can make various corresponding changes and deformations according to the present invention without departing from the spirit and essence of the present invention, but these corresponding Changes and deformations should belong to the scope of protection of the appended claims of the present invention.

Claims (28)

1. module terminal business flow controlling system, comprise a plurality of communication networks of forming by core net and Access Network, and at least one multimode terminal, described multimode terminal is equipped with a plurality of radio network interfaces that are complementary with communication network, it is characterized in that, in the Access Network of described communication network, comprise the network context monitoring modular; Described multimode terminal comprises device context monitoring modular, service identification module, decision engine module, switching executing module, wherein:

Described network context monitoring modular is used for the network context parameter of monitoring Access Network in period of time T, and after the result that will the monitor process statistical average processing, passes to the decision engine module;

Described device context monitoring modular is used for the device context parameter of monitoring multi-module mobile terminal in period of time T, and after the result that will the monitor process statistical average processing, passes to the decision engine module;

Described service identification module is used for perception service stream characteristic parameter, and the Business Stream characteristic parameter that obtains is fed back to the decision engine module;

Described decision engine module is used for network context parameter, device context parameter and Business Stream characteristic parameter, and Business Stream is analyzed, and sets up the mapping relations between Business Stream and the interface, for Business Stream is selected radio network interface;

Described multimode terminal also comprises switching executing module, and being used for according to the decision engine module is that the wave point that single Business Stream is selected starts the hand off signaling flow process, traffic flow assignment or switch on the wireless receiving and dispatching interface.

2. module terminal business flow controlling system according to claim 1, it is characterized in that, described multimode terminal also further comprises the execution result feedback module, be used for the implementation status of the former result of decision with to the sensitivity of parameter, feed back to the decision engine module, make a strategic decision according to feedback modifiers next time by the decision engine module.

3. module terminal business flow controlling system according to claim 1 and 2, it is characterized in that, described service identification module, also be used to adopt the history feature of machine learning method statistical analysis record traffic stream, and at dissimilar Business Streams, set up classification mechanism, improve the efficient of machine learning.

4. module terminal business flow controlling system according to claim 1 and 2 is characterized in that, described network context parameter comprises available bandwidth parameter, network delay parameter and the link signal mass parameter of Access Network.

5. module terminal business flow controlling system according to claim 4 is characterized in that, described network context monitoring module monitors network availability bandwidth and network delay parameter are by the active type measure mode, and improved bag is monitored algorithm.

6. module terminal business flow controlling system according to claim 1 and 2 is characterized in that, described device context parameter comprises wave point utilance parameter, battery electric quantity parameter and radio network interface transmitting-receiving power parameter.

7. module terminal business flow controlling system according to claim 3 is characterized in that, described Business Stream characteristic parameter comprises feature, Business Stream kind and the Business Stream of the Business Stream desirability parameter to service quality;

The history feature of described Business Stream is the history feature of the Business Stream of packet size, Mean Speed parameter.

8. module terminal business flow controlling system according to claim 2 is characterized in that, the mapping relations that described decision engine module is set up between Business Stream and the interface are to use Fuzzy Multi-Attribute Decision Making to carry out.

9. module terminal business flow controlling system according to claim 8 is characterized in that, described decision engine module is to adopt the pattern of input, output, feedback to carry out work;

Wherein, input variable is network context parameter, device context parameter, Business Stream characteristic parameter;

Output variable is the mapping relations of Business Stream and radio network interface;

Feedback quantity is to obtain from the execution result feedback module, is used for the auxiliary decision-making next time of revising.

10. module terminal business flow controlling system according to claim 9, it is characterized in that, the foundation that described decision engine module is made a strategic decision is the network cost function, described network cost function is by the network context parameter, parameters such as device context parameter and Business Stream feature are set up and are formed, and are expressed as follows:

fun _Net＝W _netpara _net+W _equippara _equip+W _flowpara _flow

Wherein, para _NetExpression network context parameter, W _NetThe weight of expression network context parameter, para _EquipThe indication equipment context parameters, W _EquipThe weight of indication equipment context parameters, para _FlowExpression Business Stream characteristic parameter, W _FlowThe weight of expression Business Stream characteristic parameter;

Described weight sum is 1, i.e. W _Net+ W _Equip+ W _Flow=1, the size of weight obtains according to posterior infromation before, and finely tunes according to the parameter of execution result feedback module feedback.

11. module terminal business flow controlling system according to claim 1, it is characterized in that, described multi-module mobile terminal is equipped with a plurality of Care-of Address, i.e. a main Care-of Address and a plurality of sub-Care-of Address, a plurality of sub-Care-of Address and main Care-of Address have corresponding relation, according to binding update messages result relatively, the Business Stream of coupling is directed on the new radio network interface.

12. module terminal business flow controlling system according to claim 1, it is characterized in that, described a plurality of communication network is the GPRS communication network, UMTS communication network, the combination of two or more in infrared network, wlan network, WiMAX communication network, the bluetooth communication network;

Described multimode terminal comprises a plurality of corresponding radio network interfaces, is respectively two or more the combination in GPRS interface, UMTS interface and infrared interface, wlan network interface, WiMAX communications network interface, the bluetooth communication network interface;

Described multimode terminal comprises two kinds or the combination of two or more Business Streams in electronic mail service stream, file downloading service stream, voice service stream and the multimedia service stream.

13. the communication network of a multimode terminal service stream control, form by core net and Access Network, it is characterized in that, in the described Access Network, comprise the network context monitoring modular, described network context monitoring modular is used in period of time T the network context parameter of monitoring Access Network, and the result that will monitor passes to multimode terminal through after the statistical average processing.

14. the communication network of multimode terminal service stream control according to claim 13, it is characterized in that, described communication network is for being the GPRS communication network, the UMTS communication network, the combination of two or more in infrared network, wlan network, WiMAX communication network, the bluetooth communication network.

15. the multimode terminal of a Business Stream control is characterized in that, comprises device context monitoring submodule, service identification module, decision engine module, switching executing module, wherein:

Described device context monitoring modular is used for the device context parameter of monitoring multi-module mobile terminal in period of time T, and after the result that will the monitor process statistical average processing, passes to the decision engine module;

Described service identification module is used for perception service stream characteristic parameter, and the Business Stream characteristic parameter that obtains is fed back to the decision engine module;

Described decision engine module is used for network context parameter, device context parameter and Business Stream characteristic parameter, and Business Stream is analyzed, and sets up the mapping relations between Business Stream and the interface, for Business Stream is selected radio network interface;

Described multimode terminal also comprises switching executing module, and being used for according to the decision engine module is that the wave point that single Business Stream is selected starts the hand off signaling flow process, traffic flow assignment or switch on the wireless receiving and dispatching interface.

16. the multimode terminal of Business Stream control according to claim 15, it is characterized in that, also further comprise the execution result feedback module, be used for the implementation status of the former result of decision with to the sensitivity of parameter, feed back to the decision engine module, make a strategic decision according to feedback modifiers next time by the decision engine module.

17. multimode terminal according to claim 15 or 16 described Business Stream controls, it is characterized in that, described service identification module, also be used to adopt the history feature of machine learning method statistical analysis record traffic stream, and at dissimilar Business Streams, set up classification mechanism, improve the efficient of machine learning.

18. the multimode terminal of Business Stream control according to claim 17, it is characterized in that, described multimode terminal comprises a plurality of corresponding radio network interfaces, is respectively two or more the combination in GPRS interface, UMTS interface and infrared interface, wlan network interface, WiMAX communications network interface, the bluetooth communication network interface;

Described multimode terminal comprises two kinds or the combination of two or more Business Streams in electronic mail service stream, file downloading service stream, voice service stream and the multimedia service stream.

19. a module terminal business flow controlling method is characterized in that, may further comprise the steps:

Steps A, in period of time T, Access Network is monitored its network context parameter, passes to multimode terminal; Multimode terminal is monitored its device context parameter, then network context parameter, the device context parameter of monitoring is carried out statistical average and handles; Simultaneously, multimode terminal perception service stream characteristic parameter, the Business Stream characteristic parameter of acquisition;

Step B, multimode terminal is analyzed Business Stream according to network context parameter, device context parameter and a plurality of parameters of Business Stream feature, sets up the mapping relations between Business Stream and the interface, for Business Stream is selected radio network interface;

Step C is that the wave point that single Business Stream is selected starts the hand off signaling flow process according to multimode terminal, a traffic flow assignment or switch on the wireless receiving and dispatching interface.

20. module terminal business flow controlling method according to claim 19 is characterized in that, also comprises the following steps:

Step D, multimode terminal is revised decision-making next time according to the implementation status of the former result of decision with to the sensitivity of parameter.

21., it is characterized in that described steps A comprises the following steps: according to claim 19 or 20 described module terminal business flow controlling methods

Steps A 1, in period of time T, the Access Network of communication network is monitored a plurality of network context parameters of its available bandwidth, network delay and link signal quality, and the result that will monitor passes to the decision engine module of multimode terminal through after the statistical average processing;

Steps A 2, in period of time T, the device context detection module of multimode terminal is monitored its wave point utilance, battery electric quantity and a plurality of device context parameters of radio network interface transmitting-receiving power, and the result that will monitor passes to the decision engine module of multimode terminal through after the statistical average processing;

Steps A 3, feature, Business Stream kind and the Business Stream of the service identification module perception new service flow of multimode terminal is to a plurality of Business Stream characteristic parameters such as desirability of service quality, adopt the history feature of machine learning method statistical analysis record traffic stream, and at dissimilar Business Streams, set up classification mechanism, the Business Stream characteristic parameter that obtains is fed back to the decision engine module of multimode terminal.

22. module terminal business flow controlling method according to claim 21, it is characterized in that, in steps A 1, network context monitoring module monitors network monitor available bandwidth and network delay parameter are by the active type measure mode, and improved bag monitors algorithm, comprises the following steps:

Steps A 11, structural exploration packet on the transmitting terminal multimode terminal, this probe data packet is divided into big bag and parcel;

Steps A 12, the transmission of transmitting terminal multimode terminal is how right to probe data packet, and the right order of transmission detection packet is that bag is all stamped timestamp and numbered with sending replacing mutual with parcel when these detection datas wrap in the transmitting terminal transmission greatly;

Steps A 13, the receiving terminal multimode terminal receives probe data packet, whenever receives a probe data packet and all it is stamped timestamp;

Steps A 14, receiving terminal multimode terminal are checked the detection data packet number that receives, and judge whether check the probe data packet order that receives conforms to transmitting terminal multimode terminal sending order; If do not conform to, probe data packet is abandoned, forward steps A 11 to, restart; If out of order phenomenon do not occur, then enter steps A 15;

Steps A 15, according to the timestamp of the probe data packet that receives, calculate respectively bag to the time interval, promptly the every pair of probe data packet arrives the time difference of receiving terminal respectively;

Steps A 16 is judged and is checked that whether the right time interval of parcel is greater than the big right time interval of bag; If parcel between interval greater than big bag between the time interval, then this measurement is invalid, returns steps A 11, restarts; Otherwise enter steps A 17;

Steps A 17 checks whether the right time interval of big bag is identical, if it is identical to wrap the right time interval greatly, then this measurement is invalid, returns steps A 11, restarts; Otherwise enter steps A 18 computing network effective bandwidths, and steps A 110 computing network time delays;

Steps A 18, the size in the time interval that bigger bag is right is taken out the less time interval;

Steps A 19, computing network effective bandwidth B=S/T _Min, wherein S is the packet size of big detection packet, T _MinBe to wrap that less time interval of centering greatly;

Steps A 110, relatively at interval just the time of reception of each packet stabs the difference of stabbing with transmitting time the delivery time of the packet of Fa Songing, finds out delivery time minimal data bag at interval, and the transmitting time of taking out this packet is stabbed and the time of reception stamp;

Steps A 111 is stabbed and the time of reception stamp according to the transmitting time of taking out, and utilizes linear programming method to reduce owing to send main frame and receive the asynchronous computing network time delay error that brings of host clock, calculates network delay relatively accurately.

23, module terminal business flow controlling method according to claim 22 is characterized in that, describedly also comprises the following steps: after steps A 111

Steps A 112 on access network, by the function interface that calling system provides, is gathered the link signal quality in real time;

Steps A 113 in the time cycle, is handled the monitoring result of the network effective bandwidth that calculates, network delay and link signal mass parameter respectively by the statistical average method, pass to multimode terminal.

24. module terminal business flow controlling method according to claim 21, it is characterized in that, in the described steps A 2, a plurality of device context parameters of device context monitoring module monitors multi-module mobile terminal, and the result that will monitor comprises the following steps: through the statistical average processing

Steps A 21 on portable terminal, by the function interface that device context monitoring modular calling system provides, is gathered equipment interface utilance, battery electric quantity and the radio network interface transmitting-receiving power parameter of portable terminal in real time;

Steps A 22 in period of time T, is handled parameters such as the equipment interface utilance of collecting, battery electric quantity and radio network interface transmitting-receiving power respectively by the statistical average method.

25. module terminal business flow controlling method as claimed in claim 21 is characterized in that, in step B, the history feature of described statistical analysis record traffic stream specifically comprises the following steps

Service identification module adopts the method for machine learning, and analytic record Business Stream kind, Business Stream feature and Business Stream be to the demand of QoS parameter, the history feature of the Business Stream of statistical analysis record data bag size, Mean Speed parameter.

26. multimode terminal service stream control method according to claim 22 is characterized in that:

Described big bag is that the IP head adds request of ICMP timestamp and response message, and the packet of additional data;

Described parcel is that the IP head adds request of ICMP timestamp and response message, without any the packet of additional data.

27., it is characterized in that among the described step C, the described mapping relations of setting up between Business Stream and the interface comprise the following steps: according to claim 19 or 20 described module terminal business flow controlling methods

Use Fuzzy Multi-Attribute Decision Making to set up rational mapping relations between Business Stream and the interface, wherein,

The foundation of described decision-making technique is the network cost function, and described network cost function is by the network context parameter, and parameters such as device context parameter and Business Stream feature are set up and formed, and can be expressed as follows:

fun _Net＝W _netpara _net+W _equippara _equip+W _flowpara _flow

Wherein, para _NetExpression network context parameter, W _NetThe weight of expression network context parameter, para _EquipThe indication equipment context parameters, W _EquipThe weight of indication equipment context parameters, para _FlowExpression Business Stream characteristic parameter, W _FlowThe weight of expression Business Stream characteristic parameter;

Described weight sum is 1, i.e. W _Net+ W _Equip+ W _Flow=1, the size of weight obtains according to posterior infromation before, and finely tunes according to the parameter of execution result feedback module feedback.

28. module terminal business flow controlling method according to claim 20 is characterized in that, among the described step D, described switching comprises the following steps:

Described multi-module mobile terminal is equipped with a plurality of Care-of Address, i.e. a main Care-of Address and a plurality of sub-Care-of Address, a plurality of sub-Care-of Address and main Care-of Address have corresponding relation, according to binding update messages result relatively, the Business Stream of coupling are directed on the new radio network interface.

Images