Research On Adaptive Resource Allocation For OFDM-Based Cooperative Relaying Systems

By transforming the frequency-selective multi-path propagation channel into single-path flat-fading channel in the frequency domain, OFDM (orthogonal frequency division multiplexing) can effectively eliminate the ISI (inter-symbol interference). OFDMA (orthogonal frequency division multiple access) can be considered as an extension of OFDM to multi-user scenarios, in which orthogonal subchannels are assigned to different destinations. OFDMA offers much freedom in the radio resource allocation, where the performance metric such as capacity, coverage, and fairness can be well traded off. In addition, by adding relay nodes in the cellular networks, the multi-hop transmission mechanism can further enlarge the system coverage and provide higher spectral efficiency. Besides, the relay node is not connected to the wired backbone network and is thus cost-effective. Hence, OFDM and cooperative relaying are considered as key technologies in the next-generation wireless communication networks. The combination of the two covers merits of both, and has broad research and application prospects.This thesis studies the resource allocation for OFDM-based cooperative relaying. First an overview of OFDM systems and radio resource allocation techniques is presented. The second chapter introduces the resource allocation in OFDM-based systems, including different objectives and the related solutions, e.g., the tradeoff between capacity and fairness in single cell environment, and the tradeoff between system capacity and cell-edge performance in multi-cell environment. The cooperative relaying is introduced in the third chapter, and two fixed relaying protocol is discussed briefly, namely AF (amplify-and-forward) and DF (decode-and-forward).The resource allocation in single-cell OFDM-based cooperative relaying systems is studied in detail in the fourth chapter. Several key topics in relayed transmission is discussed in previous works, including relay selection, optimal power allocation, and dual-hop subcarrier matching. Based on these, an optimal iterative power allocation method is derived for dual-hop peer-to-peer DF relaying based on the gradient method. Taking practical discrete spectral efficiency constraint into account, an optimal bit loading algorithm is developed for dual-hop peer-to-peer AF relaying. The direct or cooperative transmission strategy is adaptively selected in the bit loading process for each subcarrier along with subcarrier matching of the two hops, which can significantly reduce the system (the source and the relays) power consumption. Next, the resource allocation of multi-relay peer-to-peer communication systems is studied, and the optimization problem is formulated as a mixed binary integer programming problem by jointly considering relay selection, dual-hop subcarrier matching, and optimal power allocation. Greedy heuristic algorithms are proposed to decompose the original problem into two independent subproblems for efficient solution, which can greatly improve the system capacity and are very close to the performance upper bounds. For multi-user dual-hop relay systems, two PF (proportional fair) based resource allocation algorithms with subcarrier matching are studied, which can effectively exploit the multi-user diversity while maintaining certain fairness among users. The two algorithms are different in the order of the subcarrier allocation and matching process. With the same GPF (general proportional fairness, i.e., fairness metric), they both achieve certain capacity gain compared with the conventional PF scheduling method that is extended from single-hop system.The incorporation of relay nodes complicates the interference circumstance of the OFDM-based cooperative relaying in multi-cell environment, so coordination between neighboring access points is required in the resource allocation process. In the fifth chapter, first the interference scenario of the cooperative relaying system and some existing static/semi-static interference coordination schemes are introduced. Then, this thesis studies the subcarrier power allocation for capacity maximization of spatial multiplexing DF relay system under interference-limited environment. Using the arithmetic-geometric mean inequality approximation, the optimization problem is turned into a GP (geometric programming) problem and can be solved efficiently. Through a series of such approximation, the solution of the transformed problems converges to that of the original problem in an iterative manner. However, this algorithm has a relatively high complexity, so a greedy binary multicarrier power allocation method with low complexity is proposed. Simulation shows that both the two approaches have better performance compared to the traditional iterative waterfilling algorithm.