Research On Efficient Wireless Transmission Technologies Based On Network Coding

Applications with high-capacity,high bandwidth and high-speed are growing constantly,and network coding is one of the key supporting technologies to realize these communication requirements.The design of network coding schemes under communication technologies and application scenarios such as high-order modulation,nodes with multi-antenna,users with different computing capacity and device-to-device(D2D)transmission are faced with new challenges.In order to improve the performance of wireless transmission under the new wireless relay network and D2D scenarios,such as relay forwarding efficiency and transmission reliability,this dissertation studies the key technologies of efficient wireless transmission,such as denoising mapping and retransmission strategy based on network coding:(1)In order to reduce the symbol error ratio(SER)caused by relay forwarding,a heuristic nearest-neighbor denoising mapping physical layer network coding algorithm(HNDM)is designed to perform denoising mapping for the high-demensional superimposed constellation at relay in the two-way relay channels network with high-order modulation of the two sources.This algorithm consists of two core strategies:cluster maximization strategy and labeling strategy.The cluster maximization strategy selectively absorbs the constellation points around each constellation point,so that the nearest neighbor constellation points satisfying the exclusive law can be clustered into the same cluster as far as possible.The labeling strategy does many-to-one mapping of constellation points in each cluster,and the mapping needs to meet some constraints to ensure the successful decoding at receiver.Besides,the influence of phase difference and amplitude ratio on the decoding SER at relay is analyzed theoretically under asymmetric channel qualities of the two access channels.Simulation results show that the HNDM algorithm can effectively increase the minimum distance between clusters and reduce the decoding symbol error rate.Under certain channel conditions,when the two sources are modulated by 16QAM and 8PSK respectively,the SER of HNDM algorithm is 6dB higher than that of the Latin-matrix-aided clustering algorithm;when the two sources are modulated by 256QAM and QPSK respectively,the HNDM algorithm can improve the SER by 2dB compared with the closest neighbor clustering algorithm.(2)In order to suppress the interference between users,a multi-pair users two-hop exchanging physical layer network coding(MTE-PLNC)scheme is designed for the scenario of multi-pair users exchanging information in two-hop multi-relay wireless network.The MTE-PLNC scheme is based on the PLNC algorithm,generalized signal space alignment algorithm and zero-forcing algorithm,considering the cooperation between users,small BSs and macro BS.It consists of four transmission phases,including two multiple access phases and two broadcast phases.The constraints that the number of antennas configured by users and BSs need to meet are given;an iterative algorithm for solving alignment vector and precoding matrix is presented.We also deduce the receiving rate of two kinds of users which can complete data exchange through one-hop forwarding of small BS and joint forwarding of small BS and macro BS.The theoretical analysis and simulation results show that the proposed MTE-PLNC scheme is more suitable for this system than the signal space alignment scheme,costs less transmission time slot than multi-user MIMO scheme,and achieves higher user receiving rate than zero-forcing scheme.Under certain number of users and antennas,the achievable degree of freedom obtained by the MTE-PLNC scheme is two times higher than that obtained by the zero-forcing scheme.(3)In order to alleviate the problem of packet loss in two-hop wireless erasure network,a distributed batched sparse(BATS)coding algorithm based on spatial modulation and physical layer network coding(BATS-SM-PLNC)is proposed.To reduce bit error ratio(BER)and transmission times,a distributed BATS code coding scheme using PLNC as inner code of BATS code is designed.In order to achieve a further improvement of the performance,spatial modulation is added to modulate the transmission signal at the two sources based on the distributed BATS scheme with PLNC as inner code.The conditional pairwise error probability,average pairwise error probability and BER of relay and receiver under spatial modulation and PLNC are derived.The simulation results show that the proposed BATS-SM-PLNC algorithm can achieve lower BER and higher decoding success rate than the traditional distributed fountain code algorithm,while achieving less transmission slot overhead.Under a certain network configuration,when the encoding overhead is 2.2,the decoding success rate of BAT-SM-PLNC algorithm(SNR=12dB)is 6.5%higher than that of the distributed fountain code algorithm based on spatial modulation(SNR=10dB).(4)In dynamic D2D networks,in view of users with different computing capacity under wireless erasure channel need different packet retransmission request,using non-uniform hypergraphs strong coloring to model the problem of packets selection,we can conclude that it is a NP hard problem,and then propose a low retransmission times instantly decodable network coding(LRT-IDNC)algorithm to reduce the retransmission delay.The LRT-IDNC algorithm includes two core parts:concurrent transmitters-selecting strategy and layer-cooperation IDNC strategy.The concurrent transmitters-selecting strategy selects as many non-interfering devices as possible as the sources for each transmission slot.The layer-cooperation IDNC strategty formulates network coding scheme for each retransmission,selecting packets for combination,so as to meet the differentiated needs of different devices while ensuring the priority of base-layer packets to be received by all devices.The simulation results show that the proposed LRT-IDNC algorithm can meet different packet requirements of heterogeneous terminals with different computing capacity,and can effectively reduce the delay caused by packet retransmission.When the network connectivity is 0.5,the retransmission delay of base-layer packets of LRT-IDNC algorithm is 37.2%lower than that of the point to multi-point IDNC algorithm.