Joint Multiple-Dimensional Resource Allocation Strategies In Next Generation Wireless Relay Systems

Future wireless communication system is requested to provide high data rate, large capacity and broad coverage. Cooperative relay technique is effective to combat the fading channel, enhance the capacity and extend the coverage, which has been an important candidate technology for long term evolution-advanced (LTE-A) system. However, by introducing the relay nodes, the structure of the networks becomes more complex, and the number of wireless links increases a lot; as a result, the resource is far beyond the concept of time, frequency, code, power. Actually, the specific optimization machnism in specific technology also belongs to the "multiple-dimensional resource". Therefore, how to design the joint optimization scheme with relatively low complexity is a great challenge for both academia and standardization process, which can reveal the efficiency, reliability and robustness of future relay systems.Taking the maximum capacity as the optimization criterion, this dissertation investigates the joint multiple-dimensional resource allocation schemes for two-hop relay systems from the viewpoints of both link level and system level; moreover, it focus on the research for multihop equilibrium strategies, partial spectrum reuse methods and resource scheduling schemes for orthogonal frequency division multiplex relay (OFDM-relay) systems. Besides, the novel user group strategy is investigated for multiple-input-multiple-output OFDM relay (MIMO-OFDM relay) systems. By theoretical analysis and simulation, the effectiveness and feasibility of the proposed schemes are validated.Firstly, in traditional "three-point" cooperative relaying system, the closed form of optimal power allocation and optimal bandwidth allocation is investigated and deduced for both amplify-and-forward (AF) and decode-and-forward (DF) relays; moreover, a joint power and bandwidth allocation scheme is proposed. Besides, the system capacity of DF and compress-and-forward (CF) relay is analyzed,and a joint power and frame allocation scheme proposed. Simulation results are deployed to compare the difference among the frame, bandwidth and power allocation, and the two-dimensional resource optimization schemes can yield the superior performances.Secondly, the power allocation schemes across frequency domain, space domain as well as space-frequency planar domain are proposed for OFDM-relay system, and the optimal solutions with closed form of the three strategies are deduced under AF and DF mode with or without diversity scenarios. Furthermore, a two-hop subcarrier pairing strategy based on multihop equilibrium theorem is proposed in order to achieve joint optimization, which is suitable and available for different relay systems. The simulation results show that the proposed joint optimization scheme can improve the system capacity obviously, and the subcarrier pairing strategy can enhance the system performance further.Thirdly, for the purpose of improving spectrum utilization for LTE-A system, the intra-cell spectrum reuse strategies are investigated for OFDM-relay networks from the viewpoint of system level. First, the interference distribution of OFDM-relay system is studied with respect to no spectrum reuse and total spectrum reuse schemes, and the reasons for low signal-to-interference-and-noise ratio(SINR)are analyzed. Second, a novel partial spectrum reuse (PSR) scheme based on multihop equilibrium machnism is proposed; furthermore, an extended method is exploited for the scenario with multiple relays in one sector. Results show that the PSR which can not only enhance the direct users'capacity in the second slot obviously, but also improve the SINR of the last 5% users, so that the throughput can be maximized.Fourthly, the centralized, the semi-distributed, and the distributed scheduling strategies are investigated for cellular OFDM-relay networks. In order to improve the SINR of the users in starvation status,three fairness-oriented centralized schemes are proposed, namely greedy polling with starvation restrained, modified propor-tional fairness, and subcarrier pairing with multihop balance strategies. Based on the multihop equilibrium mechanism, a novel semi-distributed scheme is proposed, which aims to reduce the amount of feedback information. The semi-distributed scheduling method can achieve performance gain on both throughput and fraction of satisfied users. Besides, a distributed power allocation scheme is proposed based on non-cooperative game theory. The existence of Nash equilibrium of the proposed distributed scheme is proved, while it converges to a unique Nash equilibrium. At last, the comparisons for centralized, semi-distributed, and distributed strategies are given and analyzed.Fifthly, Combined with space time block code (STBC) method, the joint resource allocation schemes are investigated in both single relay and multiple relays scenarios for STBC-OFDM relay system. And the superior performance of joint optimization combined with subcarrier pairing scheme is also validated. Furthermore, a novel user group strategy combined with user's space feature and scheduling priority is proposed for MIMO-OFDM-relay networks. The proposed user group strategy not only decreases the complexity compared with the traditional schemes, but also achieves superior performances in terms of capacity and fairness.Finally, conclusions and future work are summarized in the end of this dissertation.