Resource Allocation For Heterogeneous Cooperative Relay Networks

Nowadays, the subscribers have access to all kinds of wireless services via the2Gand3G mobile communication networks, wireless local area networks (WLAN), Ad-hocnetworks, satellite networks, and etc. However, it is still quite difcult to satisfy variouskinds of wireless services simultaneously. The combination of heterogeneous wirelessnetworks have become an important technology to conquer these difculties. On the oth-er hand, cooperative relaying can enhance network coverage and reliability efectively,which has been widely recognized as one key method of the next generation heteroge-neous wireless networks. This thesis investigates wireless transmission technologies andresource managements of heterogeneous cooperative relay networks, and explores efec-tive work mode of future wireless networks.An important problem, which has not been resolved for a long time, is that theachievable data rate of cooperative relaying may be smaller than the capacity of source-destination direct transmissions in some scenarios. For this, I propose a multi-carrierdecode-and-forward (DF) relay strategy, which makes sure that the achievable rate ofwide-band cooperative relaying is no-smaller than the capacity of source-destination di-rect transmissions. By making use of transmission time multiplexing between the sourceand relay nodes and channel coding across the sub-carriers, the proposed relay strategycan fully exploit the benefits of shorter forwarding transmission distance. This strategydemonstrates that cooperative relaying can not only enhance the coverage and reliabilityof wireless networks, but also improve the spectrum efciency of wireless networks.Since the interference signal varies dynamically in realistic heterogeneous coopera-tive relay networks, we establish a model for interference prediction. The spectrum accessand resource allocation strategy of cooperative relay network is optimized to mitigate themutual interference among the wireless networks. Low complexity, real-time spectrumaccess and resource allocation algorithm with negligible sensing-transmission time de-lay is developed. Most computations of this algorithm are accomplished of-line, leavingonly simple tasks for real-time computations. The computational load of the source andrelay nodes is reduced dramatically. Real-time and optimal control of cognitive relaytransmissions can be achieved by this strategy. Then, the robustness of cognitive relaying with respect to spectrum sensing errorsis investigated. I discover and rigorous prove that the spectrum sensing error probabili-ty (the generalized Marcum Q function, which has been widely used in wireless com-munication and radar systems) is a log-concave function, which extends the analyticalproperties of spectrum sensing error probability. This novel log-concave property notonly resolves the joint optimization of sensing parameters and transmission parametersto improve the robustness of cognitive relaying, but also play an important role in oth-er optimization problems involving the spectrum sensing error probability. Some otherapplications of this log-concave property includes tight bounds and other analytical s-tudy of transmission error probabilities, outage probabilities and detection probabilitiesin wireless communications and radar communications.Finally, distributed resource allocation algorithm is proposed for heterogeneous co-operative relay cellular networks with multiple sources, multiple relays and multiple des-tinations. The time delay and large amount of data exchange of centralized resourceallocations can be avoided by this algorithm. The convergence and optimality of this dis-tributed resource allocation algorithm is proved rigorously without any approximation ofthe original resource allocation problem.