Energy and computationally efficient resource allocation methods for cellular relay-aided networks with system stability consideration

The increasing demand for coverage extension and power gain, along with the need for decreasing implementation costs, raised the idea of relaying cellular systems. Developing relay stations as a coverage extension and low cost mechanism has also brought up the challenge of utilizing the available network resources cooperatively between base stations and relays. The topic of resource allocation in the downlink of a relaying cellular system is studied in the current dissertation with the objective of maximizing transmission rate, encompassing system stability and managing the interference as it has not been investigated as a comprehensive allocation problem in the previous literature. We begin our study by modeling a single cell downlink transmission system with the objective to enhance the throughput of cell-edge users by employing decode-and-forward relay stations. We study the queue length evolution at each hop and propose a rate control mechanism to stabilize the considered queues. Accordingly, we propose a novel allocation model which maximizes user throughput with respect to the channel condition and the stability requirements. To solve the proposed allocation problem, we introduced optimization algorithm as well as heuristic approaches which offer low computation complexity. Next, we enhance the initial allocation method by considering a multi-cell system that accounts for more general and practical cellular networks. The multi-cell model embodies extra constraints for controlling the interference to the users of neighboring cells. We propose a different set of stability constraints which do not enquire a priori knowledge of the statistics of the arriving traffic. In an approach to improve the energy efficiency while respecting the stability and interference criteria, we also suggest an energy-conservative allocation scheme. We solve the defined allocation problems in a central controlling system. As our final contribution, we enhance the proposed multi-cell allocation model with a low overhead and distributed approach. The proposed method is based on the idea of dividing the resource allocation task between each base station and its connected relay stations. In addition, the messaging overhead for controlling inter-cell interference is minimized using the reference-station method. This distributed approach offers high degree of energy efficiency as well as more scalability in comparison to centralized schemes, when the system consists of larger number of cells and users. Since the defined problems embody multiple variables and constraints, we develop a framework to cast the joint design in the optimization form which gives rise to nonlinear and nonconvex problems. In this regard, we employ time-sharing technique to tackle the combinatorial format of the allocation problem. In addition, it is important to consider the situation that the time-shared approach is not beneficial when subcarriers are not allowed to be shared during one time-slot. To overcome this obstacle, we apply heuristic algorithms as well as convex optimization techniques to obtain exclusive subcarrier allocation schemes. To evaluate the performance of the proposed solutions, we compare them in terms of the achieved throughput, transmitted power, queue stability, feedback overhead, and computation complexity. By the means of extensive simulation scenarios as well as numerical analysis, we demonstrate the remarkable advantages of the suggested approaches. The results of the present dissertation are appealing for designing of future HetNet systems specifically when the communication latency and the energy consumption are required to be minimized