Throughput and fairness-aware dynamic network coding in wireless communication networks

Abstract

Abstract-Network coding techniques have received a lot of attention from the research community for providing reliable broadcasting in error-prone wireless networks. The most common network coding approach is segment coding, in which the packets are partitioned into segments, and linear network coding is performed inside each segment. In order to increase the throughput of network coding and decrease the decoding delay, dynamic coding schemes have been recently proposed. However, these methods incur many feedback messages. In this paper, we propose two dynamic network coding schemes that achieve the maximum throughput and reduce the number of required feedback messages. Moreover, we propose a fair dynamic network coding scheme that performs a trade-off between the throughput and the fairness in terms of decoding delay and the number of decodable packets at different destination nodes. Our simulation results show that our proposed dynamic network coding method provides the same throughput as the ARQ for Network Coding (ANC) method, with up to 90% less feedback messages. Moreover, our fair dynamic network coding can increase decoding delay fairness by about 80%.

Constant-Delay and Constant-Feedback Moving Window Network Coding for Wireless Multicast: Design and Asymptotic Analysis

“…Another low-overhead feedback scheme was proposed in [20] for ARQ-based network coding, where only the leading and tail receivers feed back messages to the transmitter. However, it was not discussed in [20] whether their scheme can achieve constant decoding delay for any number of receivers. To the extent of our knowledge, no previous scheme exists that can simultaneously guarantee constant decoding delay and constant feedback overhead as the number of receivers n grows, without sacrificing the throughput and reliability of wireless multicast.…”

“…It is assumed that on the feedback channel, the transmitter and each receiver can overhear each other, but the receivers may not overhear each other. Since all receivers are within the one-hop transmission range of the transmitter and in practice the feedback signals are usually sent at a much lower data rate than the normal data packet, similar to [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23]28], we assume that the feedback signals can be reliably detected.…”

“…In recent years, a class of incremental network coding schemes, e.g., [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24] are developed to resolve the long decoding delay of rateless codes. In these designs, the data packets participate in the coding procedure progressively.…”

“…Another low-overhead feedback scheme was proposed in [20] for ARQ-based network coding, where only the leading and tail receivers feed back messages to the transmitter. However, it was not discussed in [20] whether their scheme can achieve constant decoding delay for any number of receivers.…”

“…The second category of studies are centered on an incremental network coding design 1 , e.g., [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24], where the data packets participate in the coding procedure progressively. Therefore, the receivers that have decoded old packets can have early access to the processing of new data packets, instead of waiting for all the other receivers to decode the old packets.…”