Research On High-Performance Virtual Full-Duplex Relaying

Virtual full-duplex relaying(VFDR)emerges as an alternative of full-duplex device without complex hardware and software or additional costs.VFDR makes full use of the existing half-duplex devices to mimic a full-duplex relay with low cost and complexity,which has advantage of quick access,extended coverage,low bit error rate,low latency and seamless services.To achieve this goal,VFDR employs two half-duplex relays to work alternately,one of which receives the source's data while the other forwards the source's data received during the previous time slot.The process repeats until the all data from the source node is relayed to the destination node.VFDR technique,first proposed in 2004,has experienced the entire technological development of ‘prototype-stable-mature-application'.Currently,there are considerable numbers of qualified research papers on VFDR.However,challenges of VFDR still exist considering the different applications or scenarios.Therefore,this work focuses on improving and promoting VFDR technology and its application.Based on the features of selected application scenarios,the VFDR models are designed to provide the necessary impetus for the future implementation of wireless communication.This work also analyzes the performance of the proposed VFDR models: for each scenario,the interrelay interference suppression method is designed and the mathematical proof is given;and the theoretical results of efficiency(e.g.,achievable rate,throughput and multiplexing gain),reliability(e.g.,bit error rate,symbol error probability and diversity gain)and diversitymultiplexing tradeoff are obtained.These in-depth theoretical analysis is expected to be reference for VFDR designers and to reveal the advantages and prospects of VFDR in future wireless communication.The main contributions of this work are summarized as follows.1.The work of ‘differential VFDR in VANETs' proposes a differential VFDR scheme to solve the contradiction between ‘unstable link,inaccurate channel estimation' and ‘highspeed and reliable data demand' in VANETs.In VANET with high mobility and fast fading,CSI is inaccurate.Differential VFDR employs two half-duplex device roadside or in-car to mimic a VFDR.Based on superposition of coding and differential modulation/demodulation,differential VFDR provides high-speed full-duplex data relaying and reliable data detection for the unstable direct link with inaccurate CSI.In order to compact the interference without CSI,an efficient blind inter-relay interference cancellation technique is designed,which can enhance the robustness of differential VFDR system.We also prove that when the length of the smoothing filter is large enough,the inter-relay interference cancellation can achieve a good performance.The proposed scheme obtains reliable data detection without accurate instantaneous CSI,based on differential modulation and superposition coding.Moreover,this work discusses the advantages of differential VFDR,the practical application and the complexity of real-world implementation,and provides the decision rule,bit error probability and achievable rate in theory.The proposed system has a multiplexing gain approaching 1 and has a low bit error rate in time-variant fast fading channels or high-mobility scenarios.Finally,simulation is used to compare differential VFDR with conventional half-duplex relaying and channel estimation scheme in VANETs,which verifies that differential VFDR improves the performance of VANETs significantly.2.The work of ‘virtual full-duplex relaying without accurate CSI' proposes a special VFDR protocol-Double-Listening-Three-Path Successive Relaying(DL3PSR)to support the efficient inter-relay interference cancellation and data detection in a particular scenario without accurate CSI,which expands the scope of VFDR.In order to meet the challenge of interrelay interference cancellation in this scenario,this work proposes a blind inter-relay interference cancellation technique for special three-relay structure,of which the effectiveness has been proven.Then,robust cooperative non-coherent detection is proposed based on constellation mapping,XOR demodulation and joint decoding.Note that the joint decoding strategy is able to obtain two independent versions of the original data,which significantly improves the robustness of detection.Finally,detection threshold,bit error probability and achievable rate of DL3 PSR are theoretically derived.The simulation results verify that the DL3 PSR protocol is superior to the conventional two-path VFDR protocol in terms of error performance without CSI,and the data rate is close to full-duplex relaying.3.The work of ‘successive spectrum sharing for D2 D networks' proposes a win-win successive spectrum sharing scheme for cellular users(CU)and D2 D users(DU).This win-win situation can be achieved because: a pair of DU constitute a VFDR,which assists cellular transmission with enhanced capacity and improved diversity gain;meanwhile,DU can superimpose its own data with the CU data,and use “inter-relay interference channel” to obtain the access opportunity.As a result,successive spectrum sharing has the potential to significantly enhance the spectral efficiency,improve the proximity service,convert the”interference signal” into ”useful signal” and achieve win-win advantage for both high and low priority users.To eliminate the interference between cellular and D2 D links,we propose a hybrid complex field network coding(HCFNC)scheme for two cases: 1)both of the DUs correctly decode cellular data,and 2)only one DU correctly decodes cellular data.For each case,we derive the closed-form CU diversity-multiplexing tradeoff(DMT)and data rate,symbol error probability(SEP)for the HCFNC strategy,and the D2 D throughput.Our performance analysis and numerical results show that in Case I: 1)the cellular link achieves the 3 × 1 MISO DMT with full diversity;2)the cellular link achieves full-duplex mode even though all DUs are half-duplex;3)cellular link is interference-free from DU access due to HCFNC;and 4)the cellular priority is guaranteed and cellular users are incentivized for spectrum sharing;5)DUs gain full spectrum access opportunity.For all these reasons,Case I has the most potential realizability in practice.The performance of Case II depends much on the cellular direct-path,and thus it can be adopted appropriately when D2 D access is urgent and the cellular channel is not subject to deep fading.While Case II is a necessary complement to Case I,but if two suitable DUs exist,Case I is a win-win strategy.The three VFDR models are both consistent and progressive.1)Multiplexing gains are consistent.The multiplexing gain of the three VFDR models can be approximately 1(full-duplex multiplexing)as the frame length is large enough.2)Diversity gains are progressive.The potential diversity gain of the first model is 1.The reason is that: there is no direct link;the data of the source node can only be forwarded by one relay;and one copy of the data is finally decoded.The potential diversity gain of the second model is 2.The reason is that: there is no direct link;but the source node data can be simultaneously forwarded by two independent relay;and two copies of the data can be decoded with joint decoding.The third model constitutes a distributed 3 × 1 MISO system which has a diversity gain of3.Moreover,the first two models employ dedicate relays,while third model extends the scope from dedicate relay sets to mobile user sets.With the concept extension of relay sets in the third model,data transmission demand in the network includes both the relays' data exchange and the source-destination transmission.