Resource Allocation In Ofdm Based Wireless Relay Networks

The combination of relay transmission with orthogonal frequency division multiplexing (OFDM) technique is deemed as the candidate for the fourth generation (4G) wireless networks. This thesis addresses the resource allocation problem in OFDM based relay networks. Different scenarios of relay networks are considered, e.g., the multi-user relay networks, the multi-relay networks, and the cognitive radio (CR) relay networks. The resource management strategies are developed to analyse: how to optimally distribute available resources among different users, how resource optimization enhances the performance under cooperative communication, how to maximize the lifetime of multi-hop wireless sensor networks (WSNs), and how to improve CR transmission without degrading the performance of the primary network? For each relay transmission scenario, two or more resources are optimized to enhance the system performance under various constraints. The resources include: the power allocation at the source terminal, the power allocation at the relay nodes, the sub-carrier allocation among different users, and the sub-carrier matching over different hops. Multi-user transmission adopts orthogonal frequency division multiple access (OFDMA) technique and is subject to separate power constraint at each terminal. Initially, the resource management issue in multi-user uplink relay transmission is discussed. Then, the similar resource allocation problem is solved when OFDMA multi-user system operates under the bidirectional relay transmission. The multi-relay dual hop transmission is optimized under different transmission schemes, specifically, the non-orthogonal transmission where all the terminals transmit simultaneously in the second time slot and the time division multiple access (TDMA) based transmission where each relay transmits in the pre-defined time slot. The resource allocation in CR relay networks is considered which ensures that the interference caused by the secondary network is not harmful to the primary system. Convex optimization techniques are exploited to solve the problems for each relay transmission and the near optimal solutions are developed. Further, several suboptimal algorithms are also designed to reduce the computational complexity.