KR101003696B1 - A packet encoding and decoding algorithm for reducing retransmission overhead - Google Patents

A packet encoding and decoding algorithm for reducing retransmission overhead Download PDF

Info

Publication number
KR101003696B1
KR101003696B1 KR20080101891A KR20080101891A KR101003696B1 KR 101003696 B1 KR101003696 B1 KR 101003696B1 KR 20080101891 A KR20080101891 A KR 20080101891A KR 20080101891 A KR20080101891 A KR 20080101891A KR 101003696 B1 KR101003696 B1 KR 101003696B1
Authority
KR
South Korea
Prior art keywords
packet
receiver
retransmission
error
decoding
Prior art date
Application number
KR20080101891A
Other languages
Korean (ko)
Other versions
KR20100042757A (en
Inventor
김영한
김효곤
윤상기
Original Assignee
숭실대학교산학협력단
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 숭실대학교산학협력단 filed Critical 숭실대학교산학협력단
Priority to KR20080101891A priority Critical patent/KR101003696B1/en
Publication of KR20100042757A publication Critical patent/KR20100042757A/en
Application granted granted Critical
Publication of KR101003696B1 publication Critical patent/KR101003696B1/en

Links

Images

Abstract

The present invention is a retransmission method for improving the efficiency and reliability of wireless data communication. In order to overcome packet errors frequently occurring in wireless data communication, a receiver generally retransmits an error packet to a sender, which is a waste of bandwidth. This results in lower throughput. We propose a new type of retransmission method based on network coding that can overcome this problem. In the present invention, the receiver does not discard the information of the packet in error, and at the sender side, when retransmitting, the receiver does not simply transmit the previously sent packet but mixes a new packet in a network coding form. When the receiver receives the coded packet, the receiver recovers the error of the packet in which the existing error is stored. If the error is successfully recovered, the receiver decodes the other packet by using it. Unlike the conventional retransmission scheme, the present invention can reduce the retransmission overhead because the sender combines two packets and transmits the retransmission. This can be expected to improve bandwidth efficiency and throughput in wireless communications.
Network coding, retransmission, wireless communication technology

Description

Packet transmission method using encoding and decoding of packets to reduce overhead due to retransmission in wireless communication {A packet encoding and decoding algorithm for reducing retransmission overhead}

This technology is a kind of Automatic Repeat reQuest (ARQ) technique to improve the reliability of data communication. Since wireless communication is error-prone due to noise and interference in the channel, the ARQ mechanism must be used. Since ARQ retransmits the packet sent by the sender to the receiver, waste of overhead and bandwidth due to this is essential. However, the present invention proposes an ARQ scheme that can completely eliminate the overhead of retransmission according to channel conditions due to retransmission based on network coding.

Network coding is a technique for sending a plurality of packets in overlap in order to increase the efficiency of data communication. When the sender has n packets to be transmitted, the packet is mixed and then transmitted. In the case of random linear network coding, the packet is multiplied by a random coefficient and added to the packets. Use the method. In this way, the sender sends a mixed packet multiplied by a different coefficient each time. On the receiver side, if we receive n such mixed packets, we can obtain the pre-mixed packets by multiplying the mixed packets by the inverse of the coefficient matrix of the packets. The advantage of using this type of network coding is that the efficiency of retransmission is increased because only n mixed packets need to be received regardless of the type of packet. In an end-to-end packet network, the sender needs to know what packets are missing and retransmit them, so they need acknowledgment per packet. However, when network coding is used, the sender transmits packets continuously until the receiver receives n mixed packets and the receiver only needs to send an acknowledgment once, thereby reducing the waste of feedback bandwidth.

The present invention aims to reduce the overhead of retransmission by performing network coding at the code level. Conventional bit-level network coding schemes slightly increase the efficiency of retransmissions by reducing the number of ACK packets transmitted. However, if an error exists in the coded packet received by the receiver, it cannot recover and must discard the entire error packet. There are disadvantages. We propose a network coding at the symbol level that can overcome this problem.

In order to achieve the purpose, the sender transmits a packet to the receiver, the receiver transmits an ACK packet if the receiver succeeds in receiving the packet, and if it fails, it transmits a NACK packet or stores an error packet without transmitting any packet. Step 1; If it is confirmed that the receiver has failed to receive the packet, the sender transmits a mixture of two packets including information of a retransmission packet and a new packet using an encoding scheme; The receiver includes a third step of first decoding the retransmission packet using the newly received encoding packet and then decoding the new packet.

delete

delete

delete

This technique is expected to greatly reduce the bandwidth waste caused by retransmission. In general, the number of bit errors existing when an error occurs in a packet in wireless communication is often limited to only a few bits. In this situation, the combination of the retransmitted coded packet and the previously stored broken packet has a great effect and can recover a bit error with a very high probability. When the bit error is recovered, the second packet is recovered without loss of decoding probability, so it can be expected to be received with high probability, and consequently, high throughput can be expected by reducing the waste of bandwidth due to retransmission.

The present technology can be utilized in an environment such as a packet-based wireless communication network, for example, a wireless LAN based on the IEEE 802.11 standard. The sender sends a packet to the receiver, and the receiver delivers an ACK packet if the packet is successfully received, a NACK packet if it fails, or no packet. When it is confirmed that the receiver has failed to receive the packet, the sender mixes the two packets with the proposed encoding scheme. The receiver keeps existing broken packets and decodes them using the newly received encoded packets. As described above, the present invention can be applied in most packet-based wireless communication environments.
The present invention proposes a new type of packet encoding and decoding technique that operates at the symbol level. Unlike the general retransmission method, the receiver receives the error packet if the receiver receives an error and keeps the error packet without discarding it and then uses it to decode the next packet. When the sender sends a packet, if the receiver does not receive the packet, the receiver extracts and stores the confidence value of each bit from the signals of the packet. In this case, the signal means a complex number at the baseband level generated after the bandpass processing at the receiver side. The confidence value is a measure of how reliable each bit the receiver has decoded. It has a range of values from -∞ to ∞, with a negative value of 1 for decoding a bit and a positive value of 0 for decoding. do. In this case, the larger the absolute value of the confidence value, the greater the reliability of the decoded bit. To measure the confidence value of each bit, we measure the distance from the position of the received signal to the decision boundary. The measurement method is illustrated in FIG. 1, which corresponds to a case in which a sender transmits a packet using BPSK modulation.
If the receiver finds that it did not receive the packet, the sender attempts to retransmit the packet. In general, it is common for the sender to retransmit the packet itself. However, in the present invention, in order to avoid waste of bandwidth due to retransmission, the retransmitted packet is transmitted with the information of the existing packet and the new packet together. In this case, transmitting together does not increase the length of the packet but rather doubles the amount of bits per signal compared to the existing packet. To this end, mixed packets use a higher level of modulation. To illustrate this well, we use an example of BPSK. For convenience of description, a packet retransmitted is referred to as α, a packet transmitted together is β, and a packet encoded with α and β sent by a sender is referred to as α + β. 2 shows a modulation scheme for α + β packets and mapping of bit information for each constellation point. In the case of BPSK, one bit per signal is transmitted. In this technique, a retransmitted packet should deliver two bits per signal. To do this, we increase the constellation points for each signal to four. Mapping of bit information for each point is as shown in FIG. In FIG. 2, s 1 , s 2 , s 3 , and s 4 denote constellation points used by the sender to transmit data, and in each box, the value of the left box is a bit of α and the value of the right box is a bit of β. Means. The sender sends the packet encoded in this way to the receiver.
The receiver receiving α + β first decodes α through this packet. This is done by using the previously stored confidence value and the confidence value for α measured in the α + β packet. The confidence value at α + β is measured by calculating the distance from the received signal to the decision boundary as before. If the previously stored confidence value is c 1 and the confidence value for α obtained from α + β is c 2 , the determination of the bit for α is made after adding c 1 and c 2 . This makes it possible to expect the bit errors of the existing broken packet to be recovered by adding the confidence value of c 2 .
If decoding of α is successful, we use β to decode β in the α + β packet. α + β has a higher rate of modulation than conventional packets, so the error is more likely to be decoded by itself. However, in the present invention, since the information already known in decoding β utilizes information about a bit of α, the decoding probability is not different from the decoding probability of a general packet. 3 illustrates such a situation. For example, if the bit of α is 0, we can recognize that the signal sent by the receiver is s 2 or s 4 . In this case, decoding is performed by using the same decision boundary as shown in the example of BPSK of FIG. 1, resulting in the same decoding probability. Similarly, when the bit of α is 1, the signal sent from s 1 or s 3 is estimated, and in this case, the decoding probability is the same as that of BPSK.

1 illustrates an example of measuring a confidence value in a signal using BPSK modulation.

2 is a method of conveying information of α and β packets in an encoded packet for retransmission.

3 illustrates a method of decoding β using α in an encoded packet.

Claims (7)

  1. delete
  2. delete
  3. delete
  4. In the packet transmission method using the encoding and decoding of the packet to reduce the overhead of retransmission in wireless communication,
    A first step (S1) of transmitting a packet to a receiver, transmitting a ACK packet if the receiver succeeds in receiving the packet, and storing an error packet without transmitting a NACK packet or transmitting a packet if the receiver fails;
    If it is confirmed that the receiver has failed to receive the packet, the sender performs a second step (S2) of mixing the two packets together with the information of the retransmission packet and the new packet by encoding scheme;
    The receiver includes a third step (S3) of first decoding the retransmission packet and then decoding the new packet using the newly received encoding packet. Packet transmission method using encoding and decoding of a.
  5. The method of claim 4, wherein
    The first step S1 extracts and stores a confidence value of each bit from signals of an error packet. The signal means a complex number at a baseband level generated after a bandpass process at a receiver side, and the confidence value. Is a value measuring whether the receiver can trust each of the decoded bits, and has a range of values from -∞ to ∞, decodes bits to 1 for negative values, 0 for positive values, and confidence. The greater the absolute value of the value, the greater the reliability of the decoded bits, and the confidence value of each bit by measuring the distance from the position of the received signal to the decision boundary. Packet transmission method using encoding and decoding of packets to reduce the error.
  6. The method of claim 4, wherein
    In the second step S2, the packet transmission is performed by doubling the amount of bits per signal compared to the existing transport packet and transmitting the encoded and decoded packet to reduce the overhead caused by retransmission in the wireless communication. Packet transmission method using a.
  7. The method according to claim 4 or 5,
    The third step (S3) first decodes the retransmission packet, extracts the confidence value of the retransmission packet by calculating the distance from the signal position to the decision boundary, and adds the previously stored confidence value to the bit error of the existing broken packet When the decoding of the retransmission packet is successfully decoded, the new packet is decoded, and the decoding is performed using the confidence valu of the new bit measured from the position of the received signal to the decision boundary. Packet transmission method using encoding and decoding of packets to reduce overhead caused by retransmission.
KR20080101891A 2008-10-17 2008-10-17 A packet encoding and decoding algorithm for reducing retransmission overhead KR101003696B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
KR20080101891A KR101003696B1 (en) 2008-10-17 2008-10-17 A packet encoding and decoding algorithm for reducing retransmission overhead

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
KR20080101891A KR101003696B1 (en) 2008-10-17 2008-10-17 A packet encoding and decoding algorithm for reducing retransmission overhead

Publications (2)

Publication Number Publication Date
KR20100042757A KR20100042757A (en) 2010-04-27
KR101003696B1 true KR101003696B1 (en) 2010-12-24

Family

ID=42218019

Family Applications (1)

Application Number Title Priority Date Filing Date
KR20080101891A KR101003696B1 (en) 2008-10-17 2008-10-17 A packet encoding and decoding algorithm for reducing retransmission overhead

Country Status (1)

Country Link
KR (1) KR101003696B1 (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040010744A1 (en) * 1997-11-13 2004-01-15 Tao Chen Method and apparatus for time efficient retransmission using symbol accumulation
JP2006311615A (en) 2000-05-17 2006-11-09 Matsushita Electric Ind Co Ltd Hybrid arq method for packet data transmission
US20070242773A1 (en) 2006-04-18 2007-10-18 Interdigital Technology Corporation Method and apparatus for implementing h-arq in a mimo wireless communication system

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040010744A1 (en) * 1997-11-13 2004-01-15 Tao Chen Method and apparatus for time efficient retransmission using symbol accumulation
JP2006311615A (en) 2000-05-17 2006-11-09 Matsushita Electric Ind Co Ltd Hybrid arq method for packet data transmission
US20070242773A1 (en) 2006-04-18 2007-10-18 Interdigital Technology Corporation Method and apparatus for implementing h-arq in a mimo wireless communication system

Also Published As

Publication number Publication date
KR20100042757A (en) 2010-04-27

Similar Documents

Publication Publication Date Title
JP3756490B2 (en) The data transmission device
JP5502855B2 (en) Bundling of ACK information in a wireless communication system
JP3660662B2 (en) Window-type stall prevention mechanism of high-speed wireless communication system
US8537675B2 (en) Loss tolerant transmission control protocol
EP1508213B1 (en) Apparatus and method for dynamically selecting an arq method
US5600663A (en) Adaptive forward error correction system
KR100464325B1 (en) Method and apparatus for transmitting/receiving for re-transmission of packet in mobile communication system
US6700867B2 (en) Method and system for reduced memory hybrid automatic repeat request
US5828677A (en) Adaptive hybrid ARQ coding schemes for slow fading channels in mobile radio systems
US8527848B2 (en) Cooperative symbol level network coding in multi-channel wireless networks
US20020080802A1 (en) Method for multimedia communication over packet channels
US20100131817A1 (en) Method of automatic repeat request (arq) in communication system
CN1285188C (en) Self-adaptive mixed automatic repeat request method and device
US9369238B2 (en) Method and arrangement in a communication system
EP1628428A1 (en) Receiver and hybrid apq communication system
EP1347593A1 (en) Transmitter, receiver, and communication method
JP2016054532A (en) Encoding and decoding of control information for wireless communication
US20030099211A1 (en) Radio communication system
RU2421918C2 (en) Protocol of automatic repeat request having multiple mechanisms of additional feedback
US20060064625A1 (en) Extended repeat request scheme for mobile communication networks
JP5721301B2 (en) Method and apparatus for asynchronous incremental redundant reception in a communication system
US7350125B2 (en) HARQ method for guaranteeing QoS in a wireless communication system
US20140301325A1 (en) Packet data transmitting method and mobile communication system using the same
JP3394528B2 (en) Packet transmission system and packet transmission method
US8381061B2 (en) System and method for retransmissions in a wireless communications system

Legal Events

Date Code Title Description
A201 Request for examination
E902 Notification of reason for refusal
E701 Decision to grant or registration of patent right
GRNT Written decision to grant
FPAY Annual fee payment

Payment date: 20131017

Year of fee payment: 4

FPAY Annual fee payment

Payment date: 20140930

Year of fee payment: 5

LAPS Lapse due to unpaid annual fee