JP2000059346A - Error compensating method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the probability of invalidating a data packet by preferentially transmitting the following input data to which transmission within low delay time is requested, when there are plural pieces of retransmission data caused by errors while the errors frequently occur on a transmission line. SOLUTION: A transmitting station stores input data by unit of a packet in a buffer while adding sequence numbers in the order of input and transmits the data in order preferentially from the data packet inputted later in response to the receiving station's request. When the transmission fault of the first transmission packet caused by the transmission error on the transmission line is recognized from the presence/absence of a reception confirmation notice from the receiving station, such a packet is managed as a retransmission packet and the second packet inputted later than the first packet is transmitted. Thus, increase of the delay of the second packet is suppressed, and even when the error frequently occurs, the probability of invalidating the second data packet is reduced. When there is a little error, the probability of sending the first data packet to the receiving station within the prescribed delay time after the second data packet is high and worsening a loss rate is prevented.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an error compensation method for compensating for a code error occurring during data transmission by retransmitting data, and more particularly to a method for accommodating a plurality of short data packets to form a frame. The present invention relates to an error compensation method suitable for performing data transmission with low delay in a system that notifies a data transmission state and performs retransmission when a code error is detected. The present invention provides, for example, wireless A
TM (Asynchronous Transfer Mode) communication system and high-speed wireless LA
It can be used in N (Local Area Network) and the like.

[0002]

2. Description of the Related Art Conventionally, as a method for compensating for data errors in a transmission path, FEC (Fward Error Correction) has been used.
n) and ARQ (Automatic Repeat Request). In the FEC, redundant bits are added to transmission data, and a code error is corrected using the redundant bits. On the other hand, in ARQ, when a code error is detected, transmission data is retransmitted.

[0003] In the FEC, since the correction capability is fixed, many redundant bits must always be added to transmission data in order to obtain sufficient characteristics when the state of the transmission path changes. Therefore, when transmission errors occur rarely, redundant bits are useless. Therefore, it is desirable to apply ARQ in order to effectively use the transmission band.

FIG. 7 shows a processing procedure of a conventional transmitting station in a data transmission system that performs error compensation by ARQ.
Here, as a data transmission system, a transmitting station, a receiving station, and a network connected to the transmitting station as shown in FIG. 6 are assumed. When a data packet to be transmitted to the receiving station is input from the network to the transmitting station, the processing of the transmitting station proceeds from step S201 to step S202 in FIG. 7, and the input data packet is accumulated in the internal buffer of the transmitting station. .

The data packets stored in the internal buffer of the transmitting station are transmitted in order from the oldest data in step S205 in response to a request from the receiving station. That is,
When there are a plurality of data packets to be transmitted, the data packets input in the order from the network to the transmitting station are transmitted preferentially. At the receiving station, for example, FIG.
When the reception of a series of predetermined data packets (one frame) is completed, the reception confirmation notification (ACK) for the data packet whose reception has been confirmed normally or the reception confirmation is normally performed. Control information, which is a non-reception notification (NAK) for a data packet that has not been transmitted or both, is returned from the receiving station to the transmitting station.

When the transmitting station receives the control information from the receiving station in step S203 of FIG. 7, the process proceeds to step S204, and grasps the state of arrival of the data packet at the receiving station according to the received control information. Then, the status is recorded and reflected in a management table (not shown) provided in the transmitting station.
For example, if the ACK does not reach the transmitting station within a predetermined time after the transmitting station transmits the data packet, the transmitting station determines that the transmission of the data packet has failed. Then, the transmitting station manages the data packet as a data packet to be retransmitted.

[0007] In step S205, since the oldest data packet is transmitted in order, the data packet to be retransmitted is transmitted with priority over the untransmitted data packet.
That is, when an untransmitted data packet and a data packet to be retransmitted are present in the transmitting station, the latter is input to the transmitting station from the network first, so that the data packet to be retransmitted has priority. .

FIG. 8 shows an example of a data packet transmission state between a transmitting station and a receiving station when the above-described conventional error compensation method using ARQ is performed. In this example,
The following situation is assumed. That is, one data bucket is transmitted in each time slot using four transmission time slots prepared for one frame. The number of input data packets per frame is three. The allowable delay time of the data packet is within two frames after the data packet is input from the network to the transmitting station. No control is performed on the upper limit of the number of retransmissions. It is also assumed that a data packet arriving at a transmitting station cannot be transmitted in a frame arriving for reception processing, but can be transmitted only from the next frame.

As described above, the transmitting station transmits the data packets that are not normally received by the receiving station in the order of the oldest data packet based on the control information from the receiving station.

Referring to FIG. 8, the first frame and the second frame
Since all data packets of the frame have failed to be transmitted, the sequence numbers transmitted in the first frame are 1 to 3
Is transmitted prior to other data packets at the timing of the third frame. Also,
In the third and subsequent frames, since no error has occurred in all data packets, from the viewpoint of normal transmission of data packets, data packets having sequence numbers 1 to 4 transmitted in the third frame and the fourth frame have the same number. All transmitted data packets of sequence numbers 5 to 8 can be regarded as valid data packets.

In practice, however, the data packets transmitted in the third frame with the sequence numbers 1 to 3 and the data packets transmitted in the fourth frame with the sequence numbers 5 and 6 correspond to the transmitting station. Cannot be reached at the receiving side within two frames (allowable delay time) after being input to the receiving station, the receiving station does not recognize it as a valid data packet and discards it.

[0012]

When the conventional error compensation method as described above is carried out, the transmitting station transmits the data packets in order from the oldest input data packet. Therefore, a large amount of data packets to be transmitted are accumulated in the buffer of the transmitting station.
Then, sending a newly arrived data packet from the network to the transmitting station is postponed.

For example, in a communication system such as a wireless ATM, it is necessary to provide a multimedia service requiring various transmission qualities. In particular, when transmitting data such as a part of audio or video, real-time communication, that is, transmission with a low delay time is required. When the conventional error compensation method is applied to a communication service requiring a low delay time, if errors frequently occur in a transmission path, the delay time increases, and the following problem occurs.

A data packet having an excessive delay time may not be regarded as meaningful data for an end user. That is, even if a certain data packet is normally received by the receiving station, if the delay time is excessive, the data packet is equivalent to a lost data packet. Therefore, the packet loss rate in the operation of the data communication system should be evaluated in consideration of the effect of excessive delay time.

That is, in an environment where transmission with a low delay time is required, if errors frequently occur in the transmission path,
As a result, the packet loss rate may be lower than the packet error rate when error compensation by retransmission is not performed. For example, in the example of FIG.
Since six data packets were transmitted and eight of them failed to be transmitted, the packet error rate was 5
0%, but among the eight data packets transmitted in the third frame and the fourth frame, only three data packets are considered to be effective from the viewpoint of delay quality. The rate consequently increases to 81%.

In other words, not only the effect of improving the packet loss rate by retransmission is not obtained, but also the characteristics are degraded by retransmission. The present invention provides an error compensation method capable of preventing a packet loss rate from being significantly deteriorated due to frequent errors when retransmission is performed for an error in an environment where transmission with a low delay time is required. The purpose is to provide.

[0017]

According to a first aspect of the present invention, there is provided an error compensation method for managing data to be transmitted in data packet units between a transmitting station and a receiving station which transmit data via a predetermined transmission path. When transmitting an untransmitted data packet corresponding to new data sequentially input to the transmitting station from the transmitting station to the receiving station, and when a data packet that has failed in transmission due to a transmission error occurring on a transmission path occurs. Is an error compensation method for improving an error rate of a data packet by retransmitting at least a part of the data packet that failed in transmission as a retransmission data packet from the transmitting station to the receiving station, wherein the untransmitted data packet and / or the retransmission When a plurality of data packets exist as data packets to be transmitted, the order in which the plurality of data packets are input to the transmitting station is selected. There and transmits the data packet including a more rear data preferentially.

That is, in the error compensation method according to the first aspect, the data packet input to the transmitting station at a later time is transmitted earlier than the data packet input at an earlier time.
Therefore, when there are a plurality of data packets to be transmitted to the transmitting station, the order in which the data packets are transmitted changes as compared with the conventional method. For example, after transmitting the first data packet input to the transmitting station first, if the transmission of the first data packet fails when the second data packet is input from the network to the transmitting station, Since it is necessary to retransmit one data packet, the transmitting station needs to transmit the first data packet and the second data packet in order.

In this case, the second data packet input later is transmitted earlier than the first data packet. In this way, it is possible to suppress an increase in the delay time from when the second data packet is input to the transmitting station to when the second data packet reaches the receiving station. Therefore, even when errors frequently occur, the probability that the second data packet becomes invalid due to an increase in the delay time decreases. In addition, when errors do not occur frequently, even if the first data packet is retransmitted after the second data packet, the probability that the first data packet reaches the receiving station within a predetermined delay time is high. Therefore, the deterioration of the packet loss rate can be improved.

Further, for example, when there is a third data packet and a fourth data packet which have both failed to transmit at the transmitting station, it is necessary to retransmit both the third data packet and the fourth data packet. is there. Again, in this case,
As for the transmission order of the third data packet and the fourth data packet, the data packet input later from the network to the transmitting station has priority.

As a result, the delay time of the data packet to be retransmitted first is reduced, and the probability that the data packet reaches the receiving station as a valid data packet is increased. Therefore, the deterioration of the packet loss rate can be improved. Further, even if the transmission band of the transmission path is insufficient, the delay characteristics are not degraded by retransmission. In the error compensation method according to claim 1, the transmission station inputs the delay characteristic. A sequence number corresponding to the order of data is generated, each data packet is managed by the sequence number, and a priority of a data packet to be transmitted is determined by a sequence number associated with each data packet. And

According to the second aspect, since the sequence number is generated, the order in which each data packet is input to the transmitting station can be grasped by the sequence number assigned to it. With this sequence number, the priority of the data packet to be transmitted can be easily determined. For example, when assigning a sequentially increasing sequence number to a data packet sequentially input to the transmitting station, a data packet having the largest sequence number among a plurality of untransmitted data packets or retransmission data packets to be transmitted. It is sufficient to transmit in order from.

[0023]

FIG. 1 shows an example of an embodiment of the present invention.
6 to FIG. This embodiment corresponds to claims 1 and 2. FIG. 1 is a flowchart showing a processing procedure of the transmitting station of this embodiment. FIG. 2 is a flowchart showing details of step S215 in FIG. FIG. 3 is a time chart showing the data packet transmission status of this embodiment. FIG.
5 is a flowchart illustrating an example of a processing procedure of a receiving station. FIG. 5 is a graph showing the relationship between the number of data packets input in one frame and the data packet loss rate. FIG. 6 is a block diagram illustrating a configuration example of the data transmission system.

In this embodiment, it is assumed that a data packet is transmitted between a transmitting station and a receiving station as shown in FIG. Apply the method. Data packets are sequentially input to the transmitting station from the network. As a transmission path between the transmitting station and the receiving station, a wireless transmission path or a wired transmission path is assumed.

At the transmitting station, processing as shown in FIG. 1 is performed. That is, when a data packet to be transmitted to the receiving station is input from the network to the transmitting station, the processing of the transmitting station proceeds from step S211 to step S21 in FIG.
Proceeding to 2, the input data packet is accumulated in the internal buffer of the transmitting station. In this example, every time a data packet is input to the transmitting station, a sequence number indicating the input order of the data packet is generated. Then, the generated sequence number is assigned to each input data packet and managed. Actually, a sequence number is generated using an internal counter or the like that counts up each time a data packet is input.

The data packets stored in the internal buffer of the transmitting station are transmitted in order from new data in step S215 in response to a request from the receiving station. In other words, when there are a plurality of data packets to be transmitted, the data packets input later to the transmitting station from the network are transmitted preferentially. In the receiving station, for example, the processing as shown in FIG. 4 is performed. When the reception of a predetermined series of data packets (one frame) is completed, a reception acknowledgment (ACK) for the normally confirmed data packet is received. Or non-reception notification (NA
K) or the control information that is both of them is returned from the receiving station to the transmitting station.

When the transmitting station receives the control information from the receiving station in step S213 in FIG. 1, the process proceeds to step S214, and grasps the state of arrival of the data packet at the receiving station according to the received control information. Then, the status is recorded and reflected in a management table (not shown) provided in the transmitting station.
For example, if the ACK does not reach the transmitting station within a predetermined time after the transmitting station transmits the data packet, the transmitting station determines that the transmission of the data packet has failed. Then, the transmitting station manages the data packet as a data packet to be retransmitted.

In step S215, since new data packets are transmitted in order from a later input data packet, untransmitted data packets are transmitted with higher priority than data packets to be retransmitted. In other words, when an untransmitted data packet and a data packet to be retransmitted are present at the transmitting station, the order input from the network to the transmitting station is later than the former, so that the untransmitted data packet has priority. You.

The specific contents of the processing in step S215 will be described with reference to FIG. When a data packet transmission command is input to the transmitting station, the steps from step S400 to S
Go to 401. In this example, it is assumed that a data packet transmission command appears periodically at a predetermined timing for each frame. In step S401, the number of data packets to be transmitted in the frame is preset in the internal register N. For example, when four time slots are included in each frame as shown in FIG. 3, 4 is preset in the internal register N.

In the next step S402, the last generated sequence number, that is, the value of the sequence number assigned to the last input data packet is preset in the internal register SN. The sequence number is generated in step S212 in FIG. 1 each time a new data packet is input from the network to the transmitting station, and the value increases by one each time it is generated. Therefore, the last generated sequence number is the maximum value up to that time.

In step S403, it is checked whether or not one or more data packets that need to be transmitted exist in the internal buffer of the transmitting station. If there is a data packet that needs to be transmitted, the process proceeds to step S405. If there is no data packet that needs to be transmitted, the process of FIG. 2 ends, and the process returns to step S211 of FIG. Step S40
In step 5, reference is made to the management state of the data packet to which a sequence number matching the value of the internal register SN is assigned. Then, in the next step S406, it is determined whether or not transmission of the data packet is necessary. For example, if the data packet has not been transmitted or is a data packet that needs to be retransmitted, transmission is necessary, and the process proceeds to step S407.
If the data packet has already been successfully transmitted, the process proceeds to step S404 because there is no need to transmit.

In step S407, the internal register SN
Is transmitted to the receiving station, to which a sequence number matching the value of is assigned. Then, the next step S
At 408, the value of the internal register N is reduced by one. In step S409, the value of the internal register N is compared with 0. If the value of the internal register N is 1 or more, the process proceeds to step S404. When the value of the internal register N is 0,
The process of FIG. 2 ends, and the process returns to step S211 of FIG.

In step S404, the internal register SN
Is reduced by one. Then, the process returns to step S403. Accordingly, while the processing after step S403 is repeated, the sequence number (the value of the internal register SN) indicating the data packet to be processed sequentially decreases. As described above, the order of the size of the sequence number assigned to each data packet matches the order in which the data packets are input from the network to the transmitting station. Therefore,
The order of the data packets to be processed in the processing after step S403 is opposite to the input order, and newer data packets input later are processed with higher priority than older data packets. Sent to.

In this priority order, there is no particular distinction between an untransmitted data packet and a retransmitted data packet.
Since there is a high possibility that the data packet for retransmission has been input to the transmitting station before the data packet that has not been transmitted, if there is a data packet for retransmission and an untransmitted data packet, Untransmitted data packets are transmitted before data packets for retransmission.

At the receiving station, processing as shown in FIG. 4 is performed. The operation of the receiving station will be described with reference to FIG. When the receiving station receives the data packet from the transmitting station, the process proceeds from step S300 to S301. In step S301, it is detected whether an error is included in the received data packet.

If the received data packet contains an error, the process proceeds from step S301 to step S303.
, The received data packet is discarded. If no error is included in the received data packet, the process proceeds to step S302, and the data packet is stored in the buffer as normal received data. In step S304, the sequence number assigned to the normally received data packet is referred to. In the next step S305, the reception state of the received data packet is managed based on the sequence number.

In step S306, it is determined whether or not reception of all data packets included in one frame has been completed. If one frame has not ended, step S3
From 06, the process returns to S300, and the above processing is repeated. When the reception of the data packet included in one frame is completed, the process proceeds from step S306 to S307. That is, step S307 is executed once for each frame of data transmission.

In step S307, control information including the sequence numbers of all data packets received in the one frame is returned from the receiving station to the transmitting station as a reception confirmation notification (ACK). Then, the process returns to step S300 to prepare for receiving the data packet of the next frame.
An example of a transmission state of a data packet transmitted from the transmitting station to the receiving station in this embodiment will be described with reference to FIG. The major difference between the example of FIG. 3 and the conventional example of FIG. 8 lies in the order of each data packet transmitted from the transmitting station.

That is, in the conventional example shown in FIG. 8, even in the case of retransmission, data packets are transmitted in order from the oldest data packet input first to the transmitting station. On the other hand, in the example of FIG. 3, the data is transmitted in order from a new data packet input later to the transmitting station. In the example of FIG. 3, for each data packet sequentially input from the network to the transmitting station, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 10 1
Sequence numbers 1, 12, 13, 14, and 15 are assigned.

At the time when the first frame starts, the data packets held in the internal buffer of the transmitting station are three data packets having sequence numbers 1, 2, and 3. At this point, since the last generated sequence number is 3, 3 is preset in the internal register SN in step S402 in FIG. Then, in step S407 of FIG. 2, a data packet corresponding to the value of the internal register SN is transmitted, and each time step S404 is executed, the value of the internal register SN decreases by one.

Accordingly, in the first frame of FIG. 3, at the timing of each time slot in the frame, a data packet to which the sequence number 3 is allocated first is transmitted, and then a data packet to which the sequence number 2 is allocated is transmitted. A packet is transmitted, and then a data packet to which the sequence number 1 is assigned is transmitted. In the first frame of FIG. 3, since the transmission of all data packets having the sequence numbers 3, 2, and 1 has failed, the transmitting station confirms the failure with a control signal from the receiving station.

Before the start of the second frame, all three data packets having sequence numbers 3, 2 and 1 are processed by the transmitting station as data packets requiring retransmission. In the example of FIG. 3, three new data packets having sequence numbers 4, 5, and 6 are input to the transmitting station during the first frame. Therefore, when starting the second frame, six data packets having sequence numbers 1, 2, 3, 4, 5, and 6 are present in the internal buffer of the transmitting station as data packets to be transmitted.

When the second frame is started, since the last generated sequence number is 6, step S in FIG.
At 402, 6 is preset in the internal register SN.
Then, as in the case of the first frame, new data packets that are input later to the transmitting station are transmitted with priority in order. That is, in the second frame, at the timing of each time slot in the frame, the data packet to which the sequence number 6 is assigned is transmitted first, the data packet to which the sequence number 5 is assigned is transmitted, and Is transmitted, and the data packet to which sequence number 3 is allocated is retransmitted.

In the second frame shown in FIG. 3, since the transmission of all data packets having the sequence numbers 6, 5, 4, and 3 has failed, the transmitting station confirms the failure by a control signal from the receiving station. I do. Before the start of the third frame, all six data packets having sequence numbers 6, 5, 4, 3, 2, and 1 are processed by the transmitting station as data packets requiring retransmission. In the example of FIG. 3, three new data packets having sequence numbers 7, 8, and 9 are input to the transmitting station during the second frame.

Therefore, when starting the third frame,
Nine data packets having sequence numbers 1, 2, 3, 4, 5, 6, 7, 8, and 9 are present on the internal buffer of the transmitting station as data packets that need to be transmitted. When the third frame is started, the last generated sequence number is 9, and therefore, 9 is preset in the internal register SN in step S402 in FIG. Then, as in the case of the first frame, new data packets that are input later to the transmitting station are transmitted with priority in order.

That is, in the third frame, at the timing of each time slot in the frame, a data packet to which the sequence number 9 is initially assigned is transmitted,
Next, the data packet to which the sequence number 8 is assigned is transmitted, the data packet to which the sequence number 7 is assigned is transmitted, and then the sequence number 6 is assigned.
Are retransmitted.

In the third frame of FIG. 3, all the data packets having the sequence numbers 9, 8, 7, and 6 have been successfully transmitted, so the transmitting station confirms the success by the control signal from the receiving station. I do. Specifically, an acknowledgment (ACK) including the sequence number (9, 8, 7, 6) arrives at the transmitting station as a control signal. Therefore, at the end of the third frame, the data packets that need to be retransmitted are:
There are only five data packets with sequence numbers 5, 4, 3, 2 and 1. In the example of FIG. 3, three data packets having sequence numbers 10, 11, and 12 are input to the transmitting station during the third frame.

Therefore, when starting the fourth frame,
Eight data packets having sequence numbers 1, 2, 3, 4, 5, 10, 11, and 12 exist on the internal buffer of the transmitting station as data packets to be transmitted. 4th
When a frame is started, the last generated sequence number is 12, and therefore, 12 is preset in the internal register SN in step S402 in FIG. Then, as in the case of the first frame, new data packets that are input later to the transmitting station are transmitted with priority in order.

That is, in the fourth frame, at the timing of each time slot in the frame, the data packet to which the sequence number 12 is allocated is transmitted first, and then the data packet to which the sequence number 11 is allocated is transmitted. Then, the data packet to which the sequence number 10 is assigned is transmitted, and then the data packet to which the sequence number 5 is assigned is retransmitted.

In the fourth frame shown in FIG. 3, all the data packets having the sequence numbers 12, 11, 10, and 5 have been successfully transmitted, so the transmitting station confirms the success by the control signal from the receiving station. I do. Specifically, an acknowledgment (ACK) including the sequence number (12, 11, 10, 5) arrives at the transmitting station as a control signal. However,
In the example of FIG. 3, the data packet to which the sequence number 5 is assigned has a delay time from when it is input to the transmitting station to when it reaches the receiving station exceeds a predetermined allowable value. The station (or the user side) treats it as an invalid data packet.

When FIG. 3 is compared with FIG. 8, the improvement of the packet loss rate is recognized as follows. That is, in the case of the conventional method shown in FIG. 8, the packet loss rate is 81%.
On the other hand, in the example of FIG. 3, the number of transmitted data packets is 16, and 7 out of 8 data packets (sequence numbers of 6, 7, 8, 9, 1, 1) are transmitted to the receiving side without error.
0, 11, and 12) are effective, the packet loss rate is 56%, and the characteristics are improved.

In principle, the present invention guarantees that the packet loss rate can be kept lower than the packet error rate. In addition, since retransmission of an old data packet does not hinder transmission of a new data packet, low-latency communication can always be performed. The relationship between the number of data packets input in one frame and the data packet loss rate will be described with reference to FIG. In FIG. 5, the characteristics of the embodiment of the present invention are shown as “proposed method”, and the conventional example is shown as “conventional method”. The characteristics shown in FIG. 5 are obtained assuming the following conditions.

Number of data packets accommodated in one frame: 128 [data packet / frame] Data packet error rate in link: 0.01 Control link error rate: 0 Allowable delay time: 2 frames Upper limit of number of retransmissions: 1 time Input traffic: fixed Bit rate (CBR) signal The reason why the allowable delay time is set to two frames is to consider a service that particularly emphasizes real-time properties.

In FIG. 5, the horizontal axis represents the number of data packets input in one frame [packets / frame] as the speed of the input signal, and the vertical axis represents the effective packet loss rate. Referring to FIG. 5, the "conventional method" shows relatively excellent packet loss rate characteristics when the input speed is equal to or less than a certain value (that is, when a band serving as a retransmission margin is sufficient), but exceeds a certain region. It can be seen that the packet loss rate sharply rises and deteriorates beyond the packet error rate when retransmission is not performed.

On the other hand, in the case of the present invention shown in FIG. 5 as the "proposed scheme", the packet loss rate does not exceed the packet error rate even if the input speed is increased. Furthermore, if there is sufficient retransmission bandwidth margin (ie,
When the input speed is sufficiently low), a packet loss rate equivalent to that of the “conventional method” can be realized. Therefore, it can be concluded that the error compensation method of the present invention is superior to the conventional method when applied to a real-time service requiring low delay.

[0056]

As described in detail above, according to the present invention, even for various input speeds and packet error rates,
It is possible to realize a packet loss rate with a low delay and a packet error rate or less, while improving the packet loss rate according to a margin of a retransmission band. Therefore, according to the present invention, error correction can be effectively performed in real-time communication requiring low delay.

[Brief description of the drawings]

FIG. 1 is a flowchart showing a processing procedure of a transmitting station according to an embodiment.

FIG. 2 is a flowchart showing details of step S215 in FIG. 1;

FIG. 3 is a time chart illustrating a data packet transmission state according to the embodiment;

FIG. 4 is a flowchart illustrating an example of a processing procedure of a receiving station.

FIG. 5 is a graph showing a relationship between the number of data packets input in one frame and a data packet loss rate.

FIG. 6 is a block diagram illustrating a configuration example of a data transmission system.

FIG. 7 is a flowchart showing a processing procedure of a conventional transmitting station.

FIG. 8 is a time chart showing a data packet transmission state of a conventional example.

[Explanation of symbols]

Claims (2)

[Claims] 1. A method for managing data to be transmitted in units of data packets between a transmitting station and a receiving station that transmit data via a predetermined transmission path, and corresponding to new data sequentially input to the transmitting station. And transmitting an untransmitted data packet from the transmitting station to the receiving station, and retransmitting at least a part of the data packet that has failed in transmission when a data packet that has failed to transmit due to a transmission error occurring on a transmission path occurs. An error compensation method for improving the error rate of a data packet by retransmitting the data packet from the transmitting station to the receiving station, wherein a plurality of untransmitted data packets and / or retransmitted data packets exist as data packets to be transmitted. Among the plurality of data packets, the data packet including the data that is input to the transmitting station in a later order. Error compensation method, characterized by preferentially transmitted. 2. The error compensation method according to claim 1, wherein
The transmitting station generates a sequence number corresponding to the order of the data input thereto, manages each data packet with the sequence number, and associates the priority of the data packet to be transmitted with each data packet. An error compensation method characterized in that the error is determined by a sequence number.