JP2000165440A - Priority transfer control device - Google Patents

Abstract

(57) [Summary] [Object] The present invention aims to improve transfer quality by correcting transfer delay time by preferentially controlling transfer of a delayed data packet. The relay of a setup packet holds the determined start relay required time in its own device corresponding to the flow of the packet and sends the packet, and the relay of the reverse setup packet corresponds to the flow. Obtain the expected start relay time required value from the product of the start relay required time held in the own device and the time distribution coefficient set in the backward packet, the maximum allowable delay time and the expected start relay required time value, Is determined as an expected value of the required relay terminal time, and the own device holds the expected value of the relay terminal required time, the time distribution coefficient, and the maximum allowable delay time in the own device corresponding to the flow, and stores the reverse packet. The transmission and relay of the data packet are configured to transmit the delayed packet detected from the data packet with priority.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a priority transfer control device for guaranteeing a maximum allowable delay time of packet transfer between end-to-end in a network.

In recent years, applications that operate on personal computers and support multimedia such as graphics, still images, sounds, and moving images have become widespread. When such an application transmits and receives information via a network, a demand for a delay time of a packet for transferring information such as voice and moving image is severe. As described above, it is desired to provide a transmission system that performs priority transfer control of a packet so as to be within the maximum allowable delay time according to a request from an application or the like.

[0003]

2. Description of the Related Art In recent years, such service quality (Quality
of Service) Assurance methods are being actively studied. A typical method is RSVP (resource ReSerVation Pr
otocol), etc., on all the relay devices passing through the transfer path for each end-to-end packet flow (hereinafter referred to as a flow, and the flow includes a set of flows). WFQ (Weighted Fair Queuing), WRR (Weighted Round)
Robin) and other packet-scheduling techniques ensure a fixed, pre-configured bandwidth,
A technique for performing flow control is known. In such a method, it is necessary to provide a mechanism for guaranteeing the bandwidth reserved by all the relay devices on the transmission path through which the packet passes.

[0004] However, if there is a relay device without such a mechanism on a transmission path, there remains a problem that the following packet transfer delay time cannot be solved. The problem will be described with reference to FIG.

FIG. 7 is a diagram for explaining the occurrence of packet transfer delay. In the figure, 10, 20, and 30 are relay devices. Among them, the relay device 10 is a device that does not have a band guarantee mechanism, and 1, 2, 3, and 4 show flows. In the relay device 10, in a state where the packet of the flow 1 is input to the relay device 10 in a burst manner and output to another relay device (not shown), the transfer output of the packet of the flow 2 is delayed and the relay device 30 Shows the state that is input to the. On the other hand, in the relay device 20, the packet of the flow 3 is input and transferred and output, but since the absolute number of packets input to the relay device 20 is small, the flow is
The packet No. 3 is input to the relay device 30 with a short delay time.

Therefore, the relay device 30 guarantees the packet transfer amount per unit time of the input flow, but cannot perform priority control of the transfer output according to the delay time of the input packet. was there.

[0007]

That is, in the current method, since information on the maximum allowable delay time is not set in each packet for the flow 2 of the delayed packet from the relay device 10, the delayed packet is transferred to each relay. The transfer cannot be preferentially controlled by the device. In other words, there is a problem that it is impossible to recover the delay of a packet that has been delayed beyond the maximum allowable delay time.

[0008]

Means for solving the above problems will be described below.

According to a first aspect of the present invention, when a setup packet is transmitted in a priority transfer control device for transmitting a packet to a desired destination on a network, the setup packet includes a transmission reference time and is transmitted. ,
A priority transfer control device characterized in that when transmitting a data packet, the data packet is transmitted including an arrival reference time identifier.

[0010] The invention according to claim 2, wherein in the setup packet, the transmission reference time is a departure time.

The invention according to claim 3 is the priority transfer control device, wherein the setup packet includes a timeout identifier in the setup packet.

The invention according to claim 4 is the priority transfer control device, wherein the timeout identifier is a timeout period from the transmission reference time.

The invention according to claim 5 is the priority transfer control device, wherein in the data packet, the arrival reference time identifier is an arrival time limit.

[0014] The invention according to claim 6 is the priority transfer control device, wherein in the data packet, the arrival reference time identifier includes the transmission reference time and a maximum allowable delay time.

The invention according to claim 7 is the priority transfer control device, wherein the data packet includes a time distribution coefficient in the data packet.

The invention according to claim 8 is the priority transfer control device, characterized in that the data packet includes, in the data pad, information indicating whether the data packet is to be subjected to priority transfer control.

According to a ninth aspect of the present invention, in a priority transfer control device for relaying a packet received on a network to a desired destination, when a received setup packet is relayed, a start relaying requirement calculated by calculation is calculated. When the time is held in the own device corresponding to the flow of the setup packet and the setup packet is transmitted and the received reverse setup packet is relayed, the time is held in the own device corresponding to the flow. An expected start relay time required value is obtained from a product of the start relay required time and a time distribution coefficient set in the reverse setup packet, and a maximum allowable delay time and the start relay required time set in the reverse setup packet are determined. The difference from the expected value is obtained as the expected value of the required relay termination time, and the self-device corresponds to the flow to determine the required time of the relay termination. When transmitting the reverse setup packet while holding the value, the time distribution coefficient and the maximum allowable delay time, respectively, and relaying the received data packet, detecting the delayed data packet from the received data packet, A priority transfer control device characterized by transmitting the delayed data packet with priority and transmitting data packets other than the delayed data packet normally.

The invention according to claim 10 is characterized in that, in relaying the data packet, a priority transfer control device preferentially transfers a delay data packet having a large delay time from among the received delay data packets. It is.

In the invention described in claim 11, the expected start relay time required value, when relaying the setup packet, includes a current reference time of the own apparatus and a transmission reference time set in the setup packet. Priority transfer control, characterized in that a required start-point relay time is calculated from the difference between the two, and a required start-point relay required time is calculated by a product of the required start-point relay time and the time distribution coefficient held corresponding to the flow. apparatus.

In the twelfth aspect of the present invention, the expected relay end time required value is the maximum allowable delay time held in the own device corresponding to the flow and the expected start relay required time value. This is a priority transfer control device characterized by being obtained from the difference.

According to a thirteenth aspect of the present invention, the first relay end required time is calculated based on a difference between the maximum allowable delay time previously held in the own device corresponding to the flow and the expected start relay required time. Determining a time expected value, and delaying the data packet if the first expected relay termination required time is smaller than the expected expected second relay termination required time previously held in the own device corresponding to the flow. A priority transfer control device, comprising: transfer control means for performing priority transfer control as a data packet and performing normal transfer control when the data packet is not the delayed data packet.

According to a fourteenth aspect of the present invention, the required start relay time is obtained from the difference between the current reference time of the own device and the transmission reference time set in the received data packet. A first start relay required time expected value is obtained from a product of the time distribution coefficient held in the own device or a time distribution coefficient set in the data packet corresponding to the flow of the data packet, and the first start relay When the required time expected value is larger than the second required start-point relay required time value held in the own device corresponding to the flow of the data packet, it is detected as the delayed data packet and is set as a target of priority transfer control. This is a priority transfer control device characterized by the following.

According to a fifteenth aspect of the present invention, the start relay required time is obtained from the difference between the current reference time of the own device and the transmission reference time set in the received data packet. The expected value of the required start-point relay time is calculated from the product of the time distribution coefficient held in the own device or the time distribution coefficient set in the data packet corresponding to the flow of the data packet, and the transmission set in the data packet is determined. A relay expected transmission reference time is obtained from the sum of the reference time and the expected start relay time required value. When the relay expected transmission reference time is ahead of the current reference time, the data packet is preferentially transferred as the delayed data packet. This is a priority transfer control device characterized by performing control.

[0024] The invention according to claim 16 is for determining a starting relay required time from a difference between a current reference time of the apparatus itself and a transmission reference time set in the received data packet, and the starting relay required time is added to the starting relay required time. An expected start relay time required value is obtained from a product of the time distribution coefficient held in the own device or the time distribution coefficient set in the data packet corresponding to the flow of the packet, and the arrival criterion set in the data packet is calculated. A relay end required time expected value is determined from a difference between the time identifier and the start relay required time expected value. If the relay end required time expected value is ahead of a current reference time, the data packet is regarded as the delayed data packet. A priority transfer control device characterized by performing priority transfer control.

The invention according to claim 17 is for obtaining a timeout time of a sum of a transmission reference time and a timeout time set in the data packet or a timeout time set in the data packet. A priority transfer control device characterized in that the data packet is discarded when the time is later than the current reference time.

According to the present invention, in a priority transfer control device for receiving a packet from a desired source on a network, when a data packet addressed to the own device is received, the data packet is transmitted to an upper layer. However, when a setup packet addressed to the own device is received, the required start and end times are obtained from the difference between the current reference time of the own device and the transmission reference time set in the setup packet, and the specified maximum allowable delay time is determined. Is divided by the required start and end times to obtain a time distribution coefficient, generate a reverse setup packet including the time distribution coefficient, and return it to the start apparatus.

The invention according to claim 19 is the priority transfer control device, characterized in that, in the return of the reverse setup packet, the reverse setup packet includes the maximum allowable delay time.

According to a twentieth aspect of the present invention, a required start-end time is obtained from a difference between a current reference time and a transmission reference time set in the setup packet, and the required start-end time is included in the reverse setup packet. This is a priority transfer control device characterized in that:

According to a twenty-first aspect of the present invention, in a priority transfer control device for performing priority transfer control of a packet on a network, when a transfer packet is transmitted, the transfer reference time is associated with the transmission reference time and the flow of the transfer packet. Then, a time distribution coefficient previously set in the own device is set and transmitted, and when a transfer packet is relayed, the starting point is calculated from the difference between the current reference time and the transmission reference time set in the received transfer packet. The relay start time is calculated, and a first start end relay required time expected value is obtained from a product of the start relay required time and the time distribution coefficient. A comparison is made with the expected start relay time required value. If the second expected start relay time expected value is small, the transfer packet is preferentially transmitted, and if so, the transfer packet is normally transmitted. When a transfer packet is received, the start-end required time is calculated from the difference between the current reference time and the transmission reference time set in the transfer packet, and the start-end required time is set in the generated reverse transfer packet. When the reverse transfer packet is sent back to the source of the reverse transfer packet and the reverse transfer packet is relayed, the received reverse transfer packet is transmitted to the start-end device, and when the reverse transfer packet is received, The maximum allowable delay time is divided by the required start and end times set in the reverse transfer packet to obtain a time distribution coefficient, and the time distribution coefficient is automatically calculated for the flow reverse to the flow of the reverse transfer packet. This is a priority transfer control device that is stored in a device.

According to a twenty-second aspect of the present invention, when relaying the transfer packet, a start relay required time is obtained from a difference between a current reference time and a transmission reference time set in the transfer packet, and the start relay required time is determined. A priority transfer control device characterized in that an average value of a required start-point relay time and a previous start-point relay required time is calculated, and an expected value of the required start-point relay time is calculated based on the average value.

The invention according to claim 23 is the priority transfer control device, wherein the transmission reference time is the current time of each device.

The transmission reference time, reference time, arrival reference time identifier, and timeout time used in this specification may be Japan standard time or Greenwich Mean Time. Alternatively, the transmission reference time is set to 0
A relative time may be used.

The configurations of the present invention can be combined with each other as much as possible.

By providing the above method, it is possible to solve the above-mentioned problem.

[0035]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS <Basic Principle of the Invention> First, a basic principle which helps to understand an embodiment of the present invention will be described below. FIG. 1 is a diagram for explaining the basic principle of the present invention. 4
0 is an initiating device that provides a transmission function of transmitting a packet,
50 is a relay device that provides a function of relaying the packet,
Reference numeral 60 denotes a terminal device that provides a function of receiving the packet. The starting device 40, each relay device 50, and the terminating device 60 are all the same device, but are described separately for the starting device 40, the relay device 50, and the terminating device 60 according to the functions used. In addition, it is assumed that the times of the respective clocks included in the start device 40, one or more relay devices 50, and the end device 60 are the same. In order to match the times, the time of each device can be kept the same without any problem using NTP (Network Time Protocol) which is a protocol for providing the exact same time.

In the figure, the maximum allowable delay time is set by an application program or the like, and the packets transmitted from the starting device 40 are controlled by priority in each relay device so as to be within the maximum allowable delay time. The expected start relay time required value indicates the expected value of the required time from the start device to the relay device when the data is transferred from the start device to the end device with the maximum allowable delay time. Similarly, the expected value of the required relay terminal time also indicates the expected value of the required time from the relay device to the terminal device when the transfer is performed with the maximum allowable delay time. On the other hand, the required start relay time indicates the time required for the setup packet to actually travel and measured from the start device to the relay device, and the required relay end time is also the required time measured for the actual travel from the relay device to the end device. Is shown.

In the figure, each apparatus includes a transfer control means 21 for performing priority transfer control of a packet, and an expected value setting for setting information (for example, an expected value of a required time for relay termination) required for performing a priority transfer control of a packet. Means 22 and an upper layer 23 such as a communication application. The upper layer 23 is an application or the like. In particular, an application that handles audio, images, and the like has strict requirements on the delay time of packet transfer. Therefore, such an application is typically associated with the starting device 40.
, A small value is requested for the maximum allowable delay time of a packet from when the packet arrives at the terminal device 60 via the relay device 50. On the other hand, in applications that do not require real-time properties, such as e-mail, the requirements for delay time in packet transfer are not strict. Thus, it can be said that there are various requirements for the packet delay time in the upper layer 23. Here, it is assumed that the upper layer 23 sets the maximum allowable delay time with respect to the packet transfer delay time. However, if the application omits the setting of the maximum allowable delay time, or if the upper layer 23
The expected value setting means 22 can also specify or set the maximum allowable delay time.

In the present invention, for each flow, the setup packet is actually transferred from the initiating device 40 to the terminating device 60 via each relay device 50, and the time required for the initiating and terminating operations from the initiating device 40 to the terminating device 60 is calculated. The time distribution coefficient, the expected start relaying time, and the expected relay end time, which were measured and considered in consideration of the maximum permissible delay time based on the required start and end times, were calculated for each relay device 50 and for each flow. Set to. This is called an expected value setting phase.

Next, the relay device 50 from the starting device 40
, The start relay time is actually measured by passing the setup packet, and the start relay time and the time distribution coefficient (a value obtained by dividing the maximum allowable delay time by the start terminal end time).
Is held for each relay device 50 and corresponding to each flow, and the current reference time and the data packet are set when the data packet is transferred. A delayed data packet (when the relay expected transmission reference time is earlier than the current reference time of its own device) by comparing the expected transmission reference time with the expected transmission reference time obtained by adding the expected end relay required time. For example, the relay expected transmission reference time is 18:00:01
The current reference time of the own device is 18: 00: 0
Indicating a data packet such as 0 seconds)
Each of the relay devices 40 preferentially transfers the delay data packet so that the delay time to the terminal device 60 falls within the maximum allowable delay time as much as possible. Note that the detection of the delayed data packet includes a method of comparing the expected start relay required time obtained when the setup packet is transferred with the expected start relay required time calculated when the data packet is transferred, and There is also a method of comparing the expected value of the required relay terminal time obtained when transferring the packet with the expected value of the required relay terminal time obtained when transferring the data packet.

The above-described correction of the delay time of the delay data packet is called a priority transfer control phase. Next, the operations of the expected value setting phase and the priority transfer control phase will be described in detail. <Expected Value Setting Phase> The expected value setting phase performed by the expected value setting means 22 will be described with reference to FIGS. FIG. 2 is a diagram illustrating a packet flow according to the first embodiment of this invention. The initiating device 40 generates the setup packet shown in FIG. 2, sets the transmission reference time of the own device in the transmission reference time 24 field of the setup packet, and sets the timeout time or the timeout time from the transmission reference time in the timeout identifier 25. Set. Then, the setup packet is transmitted to the terminal device 6.
Send to 0. The setup packet is relayed by one or more relay devices 50 and arrives at the termination device 60. When each of the relay devices 50 receives the setup packet, the transmission reference time 24 and the timeout identifier 2
5, the expected value setting means 22 extracts the current reference time and the transmission reference time 24 set in the setup packet.
Is calculated to obtain the required start relay time. The transfer control means 2 associates the required start relay time with the current flow.
Hold at 1.

The terminal device 60 receives the setup packet, calculates the difference between the current reference time of the terminal device itself and the transmission reference time 24 set in the setup packet, and obtains the required start and end time. Further, the maximum allowable delay time is divided by the required start and end times to obtain a time distribution coefficient. Then, a reverse setup packet is generated as a response to the setup packet, the maximum allowable delay time is set to the maximum allowable delay time 26, and the time allocation coefficient is set to the time allocation coefficient 27, respectively.
Return to 0. Each of the relay devices 50 receives the reverse setup packet from the terminal device 40, and the relay device 50 can obtain the expected value of the relay terminal required time from the relay device 50 to the terminal device 60 in the following procedure. .

(1) Receive the reverse setup packet and extract the maximum allowable delay time and time allocation coefficient.

(2) A reverse flow is obtained from the flow of the reverse setup packet, and each relay device 50 extracts the required start-point relay time already held in its own device corresponding to the reverse flow.

(3) The expected start relay time is multiplied by the time distribution coefficient to obtain an expected start relay time required value. The expected value of the required start-point relay time is held in the own device in response to the reverse flow. When a data packet is transmitted, if the data packet has the time allocation coefficient 2A and the transmission reference time and the maximum allowable delay time set in the arrival reference time identifier 28, the current reference time and the data packet are used. By multiplying the difference from the set transmission reference time by the time distribution coefficient, the expected start relay time required value can be obtained.

(4) An expected relay end required time is calculated from a difference between the maximum allowable delay time and the expected start relay required time.
The relay terminal required time expected value is held in its own device in association with the reverse flow.

(5) The setup packet and the reverse setup packet are transmitted at a predetermined time interval, and each time, the average value with the previous expected relay termination required time is calculated, and the above-mentioned expected relay termination required time is set. , The fluctuation of the delay time in the packet transfer is absorbed.

The maximum allowable delay time set by the upper layer 23 is set in the reverse setup packet,
0 has been transferred. However, as described above, the expected value setting means 22 can set the maximum allowable delay time in the setup packet and send the setup packet to the terminal device 60. The setup packet and the reverse setup packet can obtain the same result in either a single packet not containing user data or a packet containing user data and information for priority transfer control.

With the above-described method, the setting of the expected value of the required relay termination time in each relay device 50 is completed, and the first expected value setting phase is completed. This is repeated at regular time intervals. <Priority transfer phase> A method of priority transfer control of a data packet will be described below.

First, the starting device 40 sets an arrival reference time to arrive at the terminating device 60 in each data packet, and sends the data packet to the terminating device 60. The transmitted data packet is transmitted to the terminating device 60 via one or more relay devices 50.
To arrive. Each relay device 50 stores the received data packets on a first-come, first-served basis, and if the data packet indicates the priority transfer within time, the relay termination from the relay device 50 to the terminal device 60 set in the expected value setting phase. From the expected time required value, the current reference time, and the arrival deadline time set for the data packet, it is calculated whether or not the data packet arrives at the terminal device 60 by the arrival reference time. It is determined whether or not to be controlled.

(1) Priority transfer in data packet 29
If the priority transfer information is set in, priority transfer is performed, and if not, priority transfer control is not performed.

(2) Extract the flow and arrival reference time of the data packet from the received data packet.

(3) Retrieve the expected value of the required relay termination time which is held corresponding to the flow.

(4) The expected value of the relay end required time is subtracted from the arrival reference time to obtain a relay expected transmission reference time.

(5) The current reference time is compared with the expected relay transmission reference time, and if the expected relay reference time is delayed, the data packet is subjected to priority transfer control.

(6) The data packets subjected to the priority transfer control are sorted in descending order using the relay expected transmission reference time as a key, and the packets with the latest expected transmission reference time are sequentially transferred with priority. Thus, the delay time is corrected by preferentially transferring the most delayed data packet, and control is performed so that the data packet can arrive at the terminal device 60 by the arrival reference time as much as possible. <First embodiment of the present invention> In a first embodiment of the present invention,
It has a transfer control unit 21 and an expected value setting unit 22.
FIG. 4 is a diagram illustrating a packet flow according to the first embodiment of this invention. The configuration of the transfer control unit 21 will be described with reference to FIG.

Reference numeral 21a denotes a packet judging unit for judging whether the input packet is a setup packet or a data packet, 21b denotes a terminating device if the destination address of the input packet is its own device, and a relay device if it is not. A device determining unit that determines that the device is a device; 21c, a data packet receiving unit that sends out the received data packet to the upper layer 23; 21d, an expected value storage unit that sets an expected value of the relay termination required time; A time-out determination unit that determines whether the delay time has exceeded the time-out time, 21 g is a packet storage unit that stores input packets, and 21 f is a packet storage unit 21.
g, a priority control packet storage unit that performs priority transfer control of the packet stored in the packet storage unit 21h, an output determination unit that controls the order of packet output according to the situation of the expected transmission reference time of the packet storage unit 21g, A data packet transmitting unit for receiving the data of the set-up packet and transmitting it to the packet transmitting unit 21j, a packet transmitting unit 21j for transmitting the packet, and a device 22b for determining whether the destination address of the input setup packet is the own device or another address. The determination unit 22a receives the setup packet, and notifies the upper layer 23 or the setup information calculation unit 22d of the result of the setting. The setting packet receiving unit 22d calculates the information necessary for setting the reverse setup packet. Information calculation unit, 22f is upper layer 2
3 or a setup packet sending unit that receives the setup packet from the setup information calculation unit 22d and sets the expected value of the required relay termination time. The setup packet sending unit 22c sends the sending reference time set in the setup packet, the current reference time information, and the start end. The expected value of the required relay terminal time from the own relay device 50 to the terminal device 60 is calculated based on the start terminal transfer time from the device 40 to the terminal device 60 and the maximum allowable delay time, and the expected value set in the expected value storage unit 21d Calculation unit, 22
“e” is a timeout time setting unit that extracts the timeout time written in the setup packet and sets the timeout time in the timeout determination unit 21e. Next, the operation will be described along the input setup packet and data packet.

When a setup packet or a reverse setup packet is input to the packet judgment unit 21a, the packet is sent to the device judgment unit 22b, and if it is a data packet, it is sent to the device judgment unit 21b. Device determination unit 22
In b, the destination of the setup packet or the reverse setup packet is sent to the setting packet receiving unit 22a when it is the own device, and when the destination is the relay device, the expected value calculating unit 22c
Send to When the destination of the received setup packet is its own device, the setting packet receiving unit 22a passes information (maximum allowable delay time, time distribution coefficient) necessary for generating the reverse setup packet to the upper layer 23, and Then, information (maximum allowable delay time, time distribution coefficient) necessary for generating the reverse setup packet is transmitted to the setup information calculation unit 22d. The setup information calculation unit 22d receives the information, calculates the information necessary for the reverse setup packet when the device is the terminal device 60, and calculates the information required by the previous relay terminal required time when the device is the relay device 50. Calculate the average value of and reset. The setting packet transmission unit 22f sets the transmission reference time and the timeout time as it is in the setup packet from the upper layer 23, or sets a timeout time obtained by adding the timeout time to the transmission reference time, and transmits the packet to the packet transfer unit 21j.

The expected value calculation unit 22c, which has received the setup packet from the device determination unit 22b,
A transmission reference time set in the setup packet;
The required start-point relay time is obtained from the difference from the current reference time, and the required start-point relay time from the start-point device 40 to the own relay device 50 and the required relay-end time from the relay device 50 to the end device 60 according to the maximum allowable delay time. The expected value is calculated, set in the expected value storage unit 21d, and transmitted to the timeout time setting unit 22e. More specifically, a transmission reference time is set in the setup packet transmitted from the start device 40, and the relay device 50 calculates the difference between the current reference time when the setup packet is received and the transmission reference time. Calculate, and the relay device 5
The required start-point relay time up to 0 is determined, set in the expected value storage unit 21d in accordance with the flow, and the setup packet is transmitted to the terminating device 60.

The terminal device 60 receives the setup packet, calculates the difference between the current reference time and the transmission reference time, and obtains the required start and end time from the start device 40 to the terminal device 60. Then, the time distribution coefficient (the result of dividing the maximum allowable delay time by the required start end time) is calculated,
The maximum allowable delay time and the time distribution coefficient are set in the reverse setup packet, and are returned to the start device 40. Each relay device 50 multiplies the required start relay time by a time distribution coefficient to obtain an expected start relay required time value. Then, by subtracting the expected start relay time required value from the maximum allowable delay time, the expected relay end time required value from the relay device 50 to the terminal device 60 can be obtained. In addition,
Although the terminal device 60 sets the maximum allowable delay time in the reverse setup packet, the start device 40 sets the maximum allowable delay time in the setup packet, and each relay device 50 sets the expected relay terminal required time described above. It is also possible to calculate and set.

The timeout setting unit 22e extracts a timeout period from the received setup packet,
The timeout time is set in the timeout time setting unit 22e for each flow. Timeout judgment unit 2
1e monitors the status of the expected transmission reference time in the packet storage unit 21g, and compares the timeout time corresponding to each flow set in the timeout time setting unit 22e with the transfer delay time of the data packet corresponding to each flow. The packet whose delay is larger than the timeout time can be deleted from the priority control packet storage unit. This function allows a user to activate or deactivate the function by setting the switch by inserting a switch between the timeout time setting unit 22e and the timeout determination unit 21e.

The expected value storage unit 21d stores the expected value calculation unit 2
2c sets the expected value of the required relay terminal time from the relay device 50 to the terminal device 60 for each flow.

The data packet sent from the device determination unit 21b is stored in a FIFO (First In F
irst Out), the priority control packet storage unit 21f
Stores the expected transmission reference time of each data packet, sorts the values of the expected transmission reference time in ascending order, and sequentially outputs the data packets from the first data packet. When a new data packet is stored in the packet storage unit 21g, it is rearranged in ascending order according to the expected transmission reference time every time it is input, and is always sorted in ascending order.

The output judging section 21h has a function of controlling the output order of data packets according to the state of the expected transmission reference time in the priority control packet storage section 21f, and detects a data packet delayed with respect to the expected transmission reference time. Then, the data packet is output preferentially, and if there is no delayed data packet, the data packet is output in the waiting order of the data packet.

The timeout time setting unit 22e monitors the status of the expected transmission reference time in the priority control packet storage unit, compares the expected transmission reference time with the current reference time,
A packet whose transfer delay time due to the difference is larger than the timeout time is stored in the priority control packet storage unit 21f.
With the ability to delete from.

As described above, priority transfer control is performed by preferentially transmitting a delay packet based on the expected value of the required relay termination time set in each relay device 50 in the setup phase, the transmission reference time, and the arrival reference time. Can be.

FIG. 5 is a diagram for explaining a function block diagram according to the first embodiment of the present invention. The priority control packet storage unit 21f shown in FIG. 4 will be described in more detail. The priority control packet storage unit 21f includes the sort key calculation unit 21f1 described in FIG. 5, a priority control information sort queue, and a clock 21z. Priority control packet storage unit 21f
Detects a delayed data packet using the packet stored in the packet storage unit 21g as a key to the expected transmission reference time,
A sort key calculation unit for outputting the expected transmission reference time to the priority control information sort queue; a priority control information sort queue for performing sorting using the expected transmission reference time as a key;
1h is an output determining unit that controls the order of outputting data packets according to the expected transmission reference time of the packet storage unit 21g, 21j is a packet transfer unit that transfers data packets, and 21z is a clock that provides the current reference time. is there.

The sort key calculation section 21f1 reads the required start relay time stored in the expected value storage section 21d based on the corresponding flow, and multiplies the required start relay time by a time distribution coefficient to calculate the expected start relay required time. Ask for. Then, the expected start relay time required value and the storage position (pointer) of the data packet in the packet storage unit 21g are stored.
To the priority control information sort queue 21f2. The priority control information sort queue 21f2 sorts in ascending order using the expected start relay time required value as a key each time the expected start relay time required value and the storage position of the data packet in the packet storage unit 21g are received. The output determination unit 21h compares the time obtained by adding the expected start relay time required value to the transmission reference time of the data packet located at the head of the priority control information sort queue with the current reference time, and when the data packet is delayed. , The data packet to the packet transfer unit 2
1j, and deletes information on the data packet from the packet storage unit 21g and the priority control information sort queue 21f2.

According to the method described above, the delay time is corrected by preferentially transferring the delayed data packet from the data packets stored in the packet storage unit on a first-come, first-served basis, and the terminating device is corrected by the arrival deadline time as much as possible. It is possible to control so that it can arrive at. <Second Embodiment of the Present Invention> First, a basic principle which assists understanding of a second embodiment of the present invention will be described below. FIG. 3 is a diagram for explaining a packet flow according to the second embodiment of this invention. It is basically the same as FIG. 2, but is characterized in that a setup packet and a data packet are included in one packet (hereinafter, referred to as a transfer packet). 1) Setup of priority control information In the starting device 40, the relay device 50, and the terminating device 60,
Next, a method for setting up information necessary for priority transfer control of a transfer packet will be described.

The transfer packet sent from the starting device 40 is sent to the terminating device 60 including the sending reference time and time distribution coefficient (details will be described later) and user data. Then, the transfer packet is transferred from the starting device 40 to the terminating device 60 via each relay device 50, and the time required for the starting and terminating is measured for each flow. Terminating device 60
Then, the start-end required time is set in the reverse transfer packet and returned to the start-end device 40. The initiating device 40 receives the reverse transfer packet and extracts the measured transfer time. Then, the maximum allowable delay time is divided by the measured transfer time to obtain a time distribution coefficient. The time allocation coefficient and the transmission reference time of the initiating device are compared with the transmission reference time 2 of the transfer packet.
4. The time distribution coefficient is set to 2A and transmitted to the terminal device 60. Each relay device 50 that has received the transfer packet
Then, the difference between the current reference time and the transmission reference time is obtained, and the start relay time is obtained. Then, the expected start relay time is multiplied by the time distribution coefficient to obtain an expected relay required time, which is set in the transfer control means 21 in correspondence with the flow of the transfer packet. When the setting is performed, an average is calculated with the already set start end relay required time expected value, and the average is set again in the expected value storage unit 21d. The priority control information is set up by the above method. 2) Priority Transfer Control Method Next, a priority transfer control method for a transfer packet in the start device 40, each relay device 50, and the end device 60 will be described.

First, the starting device 40 sets a transmission reference time, a time distribution coefficient, user data, and the like in the transfer packet, and transmits the packet to the terminating device 60. Each relay device 50 that has received the transfer packet calculates the difference between the current reference time and the transmission reference time for the transfer packet, and calculates the delay time of the transfer packet. The delay time is compared with the expected start relay time required value, and when the value of the delay time is large, the transfer packet is subjected to priority transfer control. By performing the priority control of the delay packet in each relay device 50 as necessary, the priority transfer control can be performed so that the transfer packet can reach the terminating device 60 within the maximum allowable delay time.

FIG. 6 is a diagram for explaining a functional block diagram of the second embodiment of the present invention. First, the transfer control means 21
Will be described with reference to the drawings.

The transfer control means 21 comprises the following units.

Reference numeral 21b denotes a device judging unit for judging the input transfer packet to be a terminating device if it is the own device; otherwise, a device judging unit for judging to be a relay device. 21g is a packet storage unit for storing the transfer packets in the order of input, 21f1 is a packet stored in the packet storage unit 21g, using the expected relay reference time as a key to detect a delayed transfer packet, and Is output to the priority control information sort queue, a priority control information sort queue 21f2 performs sorting using the expected relay reference time as a key,
21h is an output judging unit that controls the order of output of the transfer packets according to the state of the expected relay transmission reference time of the packet storage unit 21g, 21j is a packet transfer unit that transfers the transfer packets, and 21z is a clock that notifies the current reference time It is. The expectation setting unit 22 includes the following units.

A setting packet receiving unit 22a receives a transfer packet and notifies the upper layer 23 of the setting result.
2d is a setup information calculation unit for calculating information for setting the required start and end times for the reverse transfer packet, 22f
Is the upper layer 23 or the setup information calculation unit 22d
A packet transmission unit for receiving information from the packet transmission unit, and setting information for setting an expected value of the required termination time to an arbitrary relay device on the communication path. The packet transmission unit 22c includes: the transmission reference time set in the transmission packet; A setting packet transmitting unit that calculates the required start-end time from the start-end device to the end device based on the current reference time in step S1 and generates a reverse transfer packet.

Next, the operation of priority transfer control of a transfer packet will be described with reference to FIG. 1) In the case of the start device, the upper layer 23 generates a transfer packet, and
It requests the transfer to 0 to the setting packet sending unit 22f. The setting packet sending unit 22f sets the required start and end times in the transfer packet, and sends the transfer packet to the packet transfer unit 21j. The packet transfer unit 21j sends the data packet to the terminating device 60. 2) In the case of a relay device When the destination of a received transfer packet or a reverse transfer packet is a relay device, the device determination unit 21b sends the transfer packet to the packet storage unit 21g. In the packet storage unit 21g, if there is an indication of the priority transfer within the transfer packet, the transmission reference time and the time distribution coefficient of the transfer packet are transmitted to the expected value calculation unit 22c. Expected value calculator 2
In step 2c, the difference between the current reference time and the transmission reference time is calculated, and the required start relay time is set in the expected value storage unit 21d in correspondence with the current flow. Sort key calculator 21f1
Then, the required start-point relay time stored in the expected value storage unit 21d is read out based on the corresponding flow, and the required start-point relay time is multiplied by a time distribution coefficient to obtain an expected start-point relay required time value. Then, it sends the expected start relay time required value and the storage position of the transfer packet in the packet storage unit 21g to the priority control information sort queue 21f2. Each time the priority control information sort queue 21f2 receives the expected start relay time required value and the storage location of the transfer packet in the packet storage unit 21g, the sort is performed in ascending order using the expected start relay required time value as a key.

The output judging unit 21h compares the current reference time with the time obtained by adding the expected start relaying time to the transmission reference time of the transfer packet located at the head of the priority control information sort queue 21f2, and determines whether the transfer packet is delayed. When you are
The transfer packet is sent to the packet transfer unit 21j, and information on the transfer packet is deleted from the packet storage unit 21g and the priority control information sort queue.

The packet transfer section 21j sends the transfer packet to the terminating device 60. 3) In the case of terminating device When the destination of the received transfer packet or reverse transfer packet is the terminating device, the device determining unit 21b sends the transfer packet to the setting packet receiving unit 22a. The setting packet receiving unit 22a sends the transfer packet or the reverse transfer packet to the upper layer 23. Further, the setting packet receiving unit 22a sends the start end termination required time to the setting packet sending unit 22f in the reverse transfer packet, triggered by the received transfer packet. Setting packet sending unit 22
f sets the required start and end times for the transfer packet and sends it to the packet transfer unit 21j. The packet transfer unit 21j sends the transfer packet to the terminating device 60.

In the above, the current reference time is clock 2
1z.

According to the above method, the packet storage unit 21g
It is possible to correct the delay time by preferentially transferring a delayed transfer packet from among the transfer packets stored in the first-come-first-served order, and to control so that the packet can reach the terminating device by the arrival deadline as much as possible. Become.

[0080]

As described above, according to the present invention, in a relay apparatus accommodating two or more lines, when packets are input in bursts from one line, packets are not transmitted from the other line. Transfer is delayed. The present invention corrects the transfer delay time as much as possible by controlling the transfer of the delayed data packet preferentially so as to fall within the maximum allowable delay time set in the flow of the transfer delay packet, and improves the transfer quality of the packet. Has the effect of increasing.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a basic principle diagram of the present invention.

FIG. 2 is a diagram illustrating a packet flow according to the first embodiment of this invention;

FIG. 3 is a view for explaining a packet flow according to a second embodiment of the present invention;

FIG. 4 is a view for explaining a functional block diagram of the first embodiment of the present invention (part 1);

FIG. 5 is a view for explaining a functional block diagram of the first embodiment of the present invention (part 2);

FIG. 6 is a diagram illustrating a functional block diagram of a second embodiment of the present invention.

FIG. 7 is a diagram for explaining occurrence of packet transfer delay;

[Explanation of symbols]

 1, 2, 3, 4 Flow 10, 20, 30 Relay device 21 Transfer control means 22 Expected value setting means 23 Upper layer 24 Transmission reference time 25 Timeout identifier 26 Maximum allowable delay time 27, 2A Time distribution coefficient 28 Arrival reference time identifier 29 Priority transfer within time 2B data 40 Starter device 50 Relay device 60 Terminator

Claims (23)

[Claims] 1. A priority transfer control device for transmitting a packet to a desired destination on a network, when transmitting a setup packet, transmits the setup packet including a transmission reference time, and transmits a data packet. In this case, the priority transfer control device transmits the data packet including the arrival reference time identifier. 2. The priority transfer control device according to claim 1, wherein the transmission reference time is a departure time. 3. The priority transfer control device according to claim 1, wherein the setup packet includes a timeout identifier. 4. The priority transfer control device according to claim 3, wherein the timeout identifier is a timeout period from the transmission reference time. 5. The priority transfer control device according to claim 1, wherein the arrival reference time identifier is an arrival time limit. 6. The priority transfer control device according to claim 5, wherein the arrival reference time identifier includes the transmission reference time and a maximum allowable delay time. 7. The priority transfer control device according to claim 1, wherein the data packet includes a time distribution coefficient. 8. The data packet according to claim 1, wherein information indicating whether the data packet is subjected to priority transfer control is included in the data pad. Priority transfer control device. 9. In a priority transfer control device for relaying a packet received on a network to a desired destination, when relaying a received setup packet, the start relay time required by calculation is calculated by using the calculated start relay time. When transmitting the setup packet held in the own apparatus corresponding to the flow, and relaying the received reverse setup packet, the start-end relay required time held in the own apparatus corresponding to the flow and From the product of the reverse setup packet and the time distribution coefficient set, the expected start relay required time is calculated, and the difference between the maximum allowable delay time set in the reverse setup packet and the expected start relay required time is calculated. The relay terminal required time expected value is obtained as an expected value of the relay terminal required time, and the relay terminal required time expected value and the time distribution And transmitting the reverse setup packet while holding the maximum allowable delay time and relaying the received data packet, detecting a delayed data packet from the received data packet, and detecting the detected delayed data packet. Priority transmission control device characterized in that data packets other than delayed data packets are normally transmitted. 10. The priority transfer control device according to claim 9, wherein, in the relay of the data packet, a delayed data packet having a longer delay time is preferentially transferred from among the received delayed data packets. 11. The expected start-point relay time required value according to claim 9, wherein, when relaying the setup packet, a difference between a current reference time of the own apparatus and a transmission reference time set in the setup packet. A priority transfer control device characterized in that a required start-point relay time is calculated from the first time, and an expected value of the required start-point relay time is calculated by a product of the required start-point relay time and the time distribution coefficient held corresponding to the flow. 12. The relay end required time expected value according to claim 9, wherein the expected value of the maximum allowable delay time held in the own device corresponding to the flow and the expected start relay required time value are calculated. A priority transfer control device characterized by: 13. The first relay terminal required time according to a difference between the maximum allowable delay time previously held in the own device corresponding to the flow and the expected start relay required time. An expected value is obtained, and when the first expected relay terminal required time is smaller than the expected expected second relay terminal required time previously held in the own device corresponding to the flow, the data packet is converted to the delay data. A priority transfer control device comprising: transfer control means for performing priority transfer control as a packet and performing normal transfer control when the packet is not the delayed data packet. 14. The start relay time required from the difference between the current reference time of the own device and the transmission reference time set in the received data packet, and the start relay time is added to the start relay time. A first start relay required time expected value is obtained from a product of the time distribution coefficient held in the own device or a time distribution coefficient set in the data packet corresponding to the flow of the packet, and the first start relay When the required time expected value is larger than the second required start-point relay required time value held in the own device corresponding to the flow of the data packet, the detected value is detected as the delayed data packet;
A priority transfer control device characterized by being subjected to priority transfer control. 15. The starting relay time required from a difference between a current reference time of the own apparatus and a transmission reference time set in the received data packet, and the starting relay time is added to the starting relay time. An expected value of the required start-point relay time is obtained from the product of the time distribution coefficient held in the own device or the time distribution coefficient set in the data packet corresponding to the flow of the packet, and the transmission standard set in the data packet is obtained. A relay expected transmission reference time is obtained from the sum of the time and the start end required relay time expected value. If the relay expected transmission reference time is ahead of the current reference time, the data packet is regarded as the delayed data packet and the priority transfer control is performed. A priority transfer control device. 16. The start relay time required from the difference between the current reference time of the own device and the transmission reference time set in the received data packet, and the start relay time is added to the start relay time. An expected start relay time required value is obtained from a product of the time distribution coefficient held in the own device or the time distribution coefficient set in the data packet corresponding to the flow of the packet, and the arrival criterion set in the data packet is calculated. A relay end required time expected value is determined from a difference between the time identifier and the start relay required time expected value. If the relay end required time expected value is ahead of a current reference time, the data packet is regarded as the delayed data packet. A priority transfer control device that performs priority transfer control. 17. The timeout time according to claim 13, wherein a sum of a transmission reference time and a timeout time set in the data packet or a timeout time set in the data packet is used. A priority transfer control device, wherein the data packet is discarded when the timeout time is later than a current reference time. 18. A priority transfer control device for receiving a packet from a desired source on a network, when receiving a data packet addressed to the own device, sends the data packet to an upper layer, and sets up the address addressed to the own device. When a packet is received, the start-end required time is obtained from the difference between the current reference time of the own device and the transmission reference time set in the setup packet, and the specified maximum allowable delay time is calculated by the start-end required time. A priority transfer control apparatus characterized in that a time allocation coefficient is obtained by dividing the time division coefficient, a reverse setup packet including the time allocation coefficient is generated, and the packet is returned to the initiating apparatus. 19. The priority transfer control device according to claim 18, wherein the reverse setup packet includes the maximum allowable delay time in the reverse setup packet. 20. The required start-end time from a difference between a current reference time and a transmission reference time set in the setup packet, and the start-end required time is included in the reverse setup packet. And a priority transfer control device. 21. A priority transfer control device for performing priority transfer control of a packet on a network. When a transfer packet is transmitted, the transfer reference time is assigned to the transfer packet in advance in accordance with the transmission reference time and the flow of the transfer packet. Set the held time distribution coefficient and transmit the packet. When relaying the transfer packet, calculate the start relay time required from the difference between the current reference time and the transmission reference time set for the received transfer packet. Obtaining a first expected start relay time required value from the product of the required start relay time and the time distribution coefficient, and obtaining a second expected start relay time required value already held in its own device corresponding to the flow. And compare with
If the second expected start relay time required is small, the transfer packet is preferentially transmitted. If so, the transfer packet is normally transmitted. If the transfer packet is received, the current reference time and the current reference time are used. The start-end required time is calculated from the difference from the transmission reference time set in the transfer packet, the start-end required time is set in the generated reverse transfer packet, and the generated reverse transfer packet is returned to the transmission source of the reverse transfer packet. When relaying the forward transfer packet, the received backward transfer packet is transmitted toward the initiating device. When the backward transfer packet is received, the maximum allowable delay time is set in the backward transfer packet. A time distribution coefficient obtained by dividing the time by the required start and end times, and holding the time distribution coefficient in its own device corresponding to a flow reverse to the flow of the reverse transfer packet. The control device. 22. The relay device according to claim 21, wherein, when relaying the transfer packet, a required start relay time is obtained from a difference between a current reference time and a transmission reference time set in the transfer packet. A priority transfer control device characterized in that an average value of the required start relay time and a previous start relay time is obtained, and an expected start relay time is calculated based on the average value. 23. Claims 1, 2, 6, 11, 14 to 1
8. The priority transfer control device according to any one of 8, 20 to 22, wherein the transmission reference time is a current reference time of each device.