Research On Radio Resource Allocation In Cooperative OFDMA Systems

Under the historical background of global informatization, high-speed multimedia and data applications in wireless communications networks are set to soar. Consequently, current and future broad wireless access systems are expected to provide multiple users with a flexible service platform that can support multi-tasking services with a large variety of stringent QoS requirements in terms of BER, transmission bandwidth, delay and packet loss rate. For its capability of exploiting both the orthogonality and multiuser diversity of Orthogonal Frequency Division Multiplexing (OFDM) subcarrier space to dynamically combating channel fluctuations and reducing cochannel interference, OFDMA is regarded as the key access technology of physical layer in the next generation broadband wireless communication networks. In addition, cooperative communication techniques can take a further step to enlarge the system coverage and improve spectrum efficiency by exploring the broadcasting nature of wireless channels and cooperation among users. Due to the harsh wirelss channels and the limited resources availiable in wireless communications networks, e.g. bandwidth and power, flexible and intelligent allocation of these resources to the users is of great importance for providing the users with the best possible QoS with the least cost.After introducing a system model for the cooperative OFDMA uplink, this dissertation first illustrates some corresponding key technologies, including physical layer channel modeling, packet scheduling, wireless user cooperative strategies and inter-cell interference cancellation, all of which are closely related to resource allocation in wireless communication systems. Then based on a protocol framework with practical concerns, the original radio resource allocation problem is formulated as a constrained nonlinear programming problem. The increment of constraint dimensions resulting from user cooperation and inter-cell interference cancellation, together with the difficult nonconvexity nature of this program, greatly complicate the process of optimal solution seeking. The formerly used optimization methods based on rigorous mathematical theory now become too complex for cost-effective real-time implementation in most practical cooperative OFDMA systems being considered for deployment, namely, timeliness requirements of system operation can not be satisfied with these methods. Based on the dual optimization and decomposition theory, the original optimization problem is analyzed in this dissertation, and the results of theoretical analysis are exploited to guide practical algorithm development. The general algorithm framework proposed in this dissertation views the problem of joint subcarrier and power allocation as a combination of a marginal rate constrained problem with equal power allocation assumption and an extra power allocation problem to reduce computational complexities. Numerical results and comparisons with various other resource allocation algorithms illustrate the fact that the proposed algorithm can obtain a near-optimal solution with comparatively low computational complexity, so that the proposed algorithm to a great extent can improve the system resource efficiency as well as system performance given that users'QoS guarantees are ensured.