Research On Two-way Wireless Relay Transmission Technology Based On Network Coding

In order to exchange information of mutual interest,two wireless terminals that are too far apart or that have severe channel fading can use a two-way wireless transmission channel formed by multiple relay devices to forward their data packets in sequence.At this point,two terminal nodes and multiple relay nodes constitute a two-way wireless relay network(TWRN).Existing research has proved that compared with the traditional Store-and-Forward(SF)method,the transmission technology based on Amplify-andForward(AF)-based Analog Network Coding(ANC)and Decode-and-Forward(DF)-based Digital Network Coding(DNC)can provide theoretical gains in transmission throughput of up to 100% and 33.3% for TWRN,respectively.Because the analog network coding will amplify the noise at the relay node,and there are technical difficulties such as high-precision clock synchronization and channel state estimation,this thesis focuses on the TWRN transmission technology based on digital network coding.In order to solve the problem that the digital network coding technology in TWRN will lead to the propagation of single transmission errors in the system,this thesis proposes a non-retransmission TWRN transmission scheme with the capability to prevent error propagation based on DNC.This scheme makes use of the feature that a single transmission error will cause the encoded packets buffered by all network nodes to have error bits at the same positions.By designing an appropriate buffering and coding strategy for each network node,after the terminal node receives and decodes an error packet for the first time,the error bits between the subsequently received erroneous packet and the buffered erroneous packet can always cancel each other,so that the correct packet can always be obtained in the subsequent receiving and decoding process.The advantages of this scheme include:(1)It can support two terminal nodes of a TWRN with any number of relays to exchange 2 data packets every 3 timeslots and hence achieve similar or better theoretical transmission throughput and spectral efficiency than the existing schemes of the same style;(2)It guarantees that a single transmission error will make two terminal nodes to decode first one pair of erroneous data packets only and then correct ones thereafter;(3)All network nodes do not need to retransmit erroneous packets,which can effectively avoid the additional overhead required for packet retransmission without reducing the transmission throughput of the TWRN.Numerical simulation shows that,in the Rayleigh fading channel environment,the proposed DF-based DNC scheme can support a TWRN at low signal-to-noise ratio(SNR)to achieve a larger effective throughput than the existing three schemes of the same style and afford a TWRN at high SNR with a network transmission throughput closely approaching that of an ideal TWRN with error-free transmission.