Research On Resource Allocation Algorithms For Wireless Relay Networks

Recently, the mobile internet has experienced an explosive development, and the massive data processing requires more efficient wireless technology urgently. Wireless relay technology has been shown to provide more reliable transmission and broader coverage for next-generation wireless networks. The resource allocation results of the relay system decide directly the resource efficiency and the system performances.The resource allocation schemes for wireless relay networks are presented in this thesis. The emphasis is laid on multi-relay selection, power allocation (PA) and subcarrier pairing (SP). The research details and the main contributions of the emphasis are as follows:(1) For amplify-and-forward (AF) based multi-relay networks, in order to minimize the system outage probability with a required transmission rate under both total and individual power constraints, a novel joint power allocation and multi-relay selection scheme is proposed. Besides the channel condition, the proposed scheme also takes relays’maximal output ability into consideration. Also, the idea of ordering is adopted to avoid exhaustive search to get the optimal relay subset with much lower complexity. In addition, simple power reallocation method is provided to avoid repetitive PA operation during the process of searching all possible relay subsets. Simulation results show that the proposed multi-relay selection scheme provides similar performance compared to the optimal scheme with optimal PA and exhaustive search but with much lower complexity.(2) A two-way AF-based multi-relay cognitive radio (CR) network is considered. A time-efficient sub-optimal power allocation scheme based on Cauchy-Schwarz inequality is provided, which is aimed to maximize the transmission rate of the secondary system and ensure the quality of service of the primary user. Simulation results show that the proposed scheme outperforms the single best relay scheme and the equal power allocation scheme, and performs very close to the near-optimal power allocation scheme using interior point method.(3) The thesis considers an orthogonal frequency division multiplexing (OFDM) based multi-relay network, and a joint optimization scheme for PA and SP is proposed to maximize the transmission rate under total power constraint. In the proposed resource allocation scheme, SP is allowed among relays and a certain subcarrier pair can be forwarded by relays to provide space diversity. Different from traditional cooperative OFDM networks, in the second phase, the source node is allowed to retransmit the same signal sent in the first phase on another subcarrier, thus improving the system performance further. The optimization problem is simplified by introducing the definition of equivalent channel gain, and a new PA scheme based on the Cauchy-Schwarz inequality is provided. Then a suboptimal scheme avoiding massive calculation of equivalent channel gain is proposed, which can provide similar performance with lower computation complexity. Simulation results show that the proposed schemes outperform other schemes with or without space diversity.(4) At last, based on the aforementioned researches, a resource allocation scheme based on SP and PA for energy-efficient communication in the cooperative OFDM network is studied under the system minimum transmission rate constraint. The energy-efficiency (EE) optimization problem is formulated into a mixed-integer nonlinear programming (MINLP) problem, which has intractable complexity due to its combinatorial nature. By defining the equivalent channel gain, the optimizing model is simplified as a typical fractional programming problem, which is proved to be quasi-concave, to solve the complex the MINLP problem. Thus Dinkelbach’s fractional programming method, subgradient method and the Hungary algorithm can be used to find the optimal solution to the proposed MINLP problem. Simulation results show that the proposed EE scheme can ensure the minimum transmission rate requirement and promote the EE of the system with rapid global convergence ability.