Resource Allocation For Virtual MIMO-based Cooperative Relaying Networks

By taking advantage of the antennas of neighboring nodes,the cooperative com-munication technique can significantly improve system capacity or transmission relia-bility,which helps the single antenna system achieve approximated performance with MIMO technology.To exploit the spatial diversity in the cooperative relaying system,the issue of relay selection and power allocation(RSPA)is of vital importance and has attracted a great attention.In this dissertation,we first study the RSPA problem in the dual-hop relaying net-works to reduce the overhead of channel state information.The objective is to minimize the outage probability assuming only the mean channel gain information is available.We propose to decompose the original RSPA problem into two tractable subproblems,which can be implemented in a semi-distributed manner.For the relay selection(RS)subproblem,a relay ordering-based scheme is designed to incrementally select a sub-set of relays according to a specific metric dominated by the mean channel gain of each individual nodes.For the power allocation(PA)subproblem,a source-driven iterative algorithm is proposed for each relay to decide its optimal transmit power locally by exploiting the convexity of this subproblem.Simulation results shown that the pro-posed scheme can dynamically select relays and adjust power allocation according to the signal-to-noise-ratio(SNR)and channel conditions.It outperforms a benchmark PA-only scheme,and the performance is close to the optimal scheme with exhaustive search,while the computational complexity is significantly reduced.Based on the study of the conventional dual-hop cooperation framework,it is inef-ficient when implemented in clustered wireless networks,since relays with imbalanced channels to the source and the destination may become the bottleneck of the overall transmission.We then focus on exploiting the spatial diversity in clustered wireless networks.A novel cooperative framework called Three-Stage Relaying(TSR)scheme is specifically proposed for networks with clustered topology.In TSR,relays are di-vided into two clusters/groups,and a virtual MIMO antenna array is formed by intro-ducing the transmission between two relay groups.TSR enables transmissions over shorter distances,which naturally breaks the bottleneck emerging in the dual-hop co-operation.An optimization problem is formulated to deal with the RS issue in TSR,aiming at maximizing the received SNR of the destination subject to the number of relays constraint.The problem is nontrivial due to the coupling of the relays in two groups.Thus,we decompose it into two subproblem,one deals with the selection of transmitting relays,the other deals with the selection of receiving relays.Two heuris-tic algorithms that achieve the tradeoff between the optimality of the solution and the computational complexity are proposed.Extensive simulations show the superiority of TSR over the dual-hop cooperation schemes in both the clustered and centered wireless networks in terms of symbol error rate and throughput.The diversity gain can be further improved by leveraging the different channel conditions of the relays and different transmit power levels among the transmitting terminals(including the source and relays)in TSR.Therefore,we continue to investi-gate the PA problem based on the TSR framework,with the objective of minimizing the outage probability under the total power constraint.To address the computation-al complexity,the problem is decomposed into two subproblems,one deals with the PA issue of the source and the first-hop relays,the other deals with the PA issue of the second-hop relays.We design algorithms for each subproblem by exploiting their special structure,and then develop a master procedure to handle the power allocation across them.The performance of the proposed scheme is evaluated through simulation,which shows that the TSR framework achieves significant improvement on the outage probability compared with the dual-hop cooperation,and the power consumption is more fairly distributed across relays.