The Research Of Differential Network Coding Technology And Detection Algorithm In TWRC Model

Network coding (NC) has already applicated in a variety of systemmodel. On the one hand, NC technology combines with cooperationcommunications (CC), analog network coding with cooperation AF anddigital network coding with cooperation DF were proposed which can furtherreduce time-slots overhead. On the other hand, space-time network codingmerges multiple input multiple output (MIMO) technology into NC also canguarantee good system performance and high transmission rate. In order toavoid channel estimation, more concerns motivate the network coding basedon non-coherent modulation. These schemes can combine diverse networkcoding strategies and detect signals depending on non-coherent detection atthe receiver. The non-correlation detection algorithms for differentialnetwork coding (D-NC) mainly include single symbol differential detection(SSDD) and multiple-symbol differential detection (MSDD). In SSDD, thedecoder detects single orginal symbol for each time. But in MSDD, thereceiver exploits N (N>2) correlative received symbols to detect N-1orginalsymbols jointly. Although the latter has higher computational complexity, theperformance becomes better. This paper adopts different optimizationstrategies and searching for a better compromise between performance andcomplexity to design more excellent MSDD algorithm in differential networkcoding.This paper considers the study of differential network coding (NC) basedon two-way relay channels (TWRC). Firstly, we propose non-coherent NCscheme which combines with different cooperation communications protocols.Secondly, we further improve previous scheme and put forward new betterscheme with the combination of MIMO technology. Meanwhile, our schemeincludes the research of multiple-symbol differential detection based onnon-coherent modulation. At last, the performance analyses of improvementscheme are completed. In this paper, section2summarizes several cooperative NC based on relay AF, DF protocols in TWRC model, whichcombine with differential relay protocol in CC and relay on correlationdetection and non-correlation detection. By Comparing time-slot usage ofthese schemes, the proposed schemes can reduce time-slot number in a largeextent when plenty of data are transmitted. It has an obvious advantage onsaving time-slot resource. Simulation results verify the performance of NCbaced on AF protocol is better than NC baced on DF protocol. Section3proposes a differential network coding scheme based on detection andforward protocol (DF-DNC) at relay node in TWRC model. At relay node andsource nodes, this work mainly research multi-symbol differential detectionmaximum likelihood (MSDD-ML) algorithms and two lower complexitysphere decoding algorithms, namely depth-first multi-symbol differentialsphere decoding (MSDSD) and width-first MSDSD to reduce computationalcomplexity. Simulation results verify the performance of two lowercomplexity algorithms is close to MSDD-ML algorithm and they complexityreduces obviously. Furthermore, width-first MSDSD has lower complexitythan depth-first MSDSD. Section4combines with MIMO technology anddigital network coding, proposes an orthogonal differential space-timenetwork coding (ODSTNC) scheme for the MIMO-TWRC system which has awider coverage and better transmission performance. The scheme can achievegood performance for power limited circumstances such as far distancetransmission between two terminals. Simulation results also have verifiedthat the ODSTNC scheme has a good performance as expected. Similary, thissection also researches MSDD in ODSTNC scheme, and analyzes the averageSER under DBPSK modulation. The upper and lower bounds on symbol errorrate (SER) have been achieved.