Resource Allocation in Autonomous Cellular Networks

With a fixed number of terminals and a fixed system spectrum, it can be shown that increasing the number of BSs in a cellular network results in a better network performance. However, a regular organized deployment of a cellular network with small cells is both prohibitively expensive and complicated. A possible solution is to allow the network to grow in an organic fashion. This calls for self-configurable approaches for resource allocation which not only improve the performance but also decrease the network planning as well as maintenance costs. Over the past few years, a particular type of small BSs known as femto-BSs has attracted tremendous attention from the industry and academia alike. Femto BSs operate on the licensed spectrum allocated to the cellular network. They are small, inexpensive and transmit at low power levels. Although the deployment of these small BSs result in significant improvements in the indoor coverage for both voice and data services, they only satisfy the immediate needs of the present cellular networks — namely the improvement of the indoor coverage. They do not provide a fundamental solution for a scalable architecture which can organically grow and adapt to the short and long term changes in the network. In this thesis, we propose a practical unified framework for the system design and resource allocation of cellular networks with various types of base stations (with different power budgets) deployed in an irregular fashion. We refer to these networks as autonomous cellular networks. To this end, we first introduce an adaptive localized approach to resource allocation in traditional cellular networks. We show that localized resource assignment at the cost of partial loss of orthogonality at the cell edges can provide significant gains in the network performance. In the second part of the thesis, we propose an adaptive self-configurable resource allocation framework for cellular networks. A network clustering technique is proposed which enables coordinated resource allocation in a network with irregular deployment of BSs. The performance of the proposed framework is evaluated with the regular as well as the irregular deployment of BSs.Ph