Abstract
In current day modern multicell networks, inter-cell interference (ICI) poses a major obstacle to achieving the required high data rates and limits the performance of mobile users with the cell-edge and heavily shadowed users, i. e. , critical users, being the worst affected. The main objective of this thesis is to address and analyse the multicell networks main problem: the ICI effect on the critical users’ performance. By offering to reduce, mitigate, avoid or even exploit the interference and through user grouping methods, this thesis presents various ICI coordination (ICIC) techniques to improve the performance of critical users, as a primary target, while maintaining the overall system performance, as a secondary target. Limited cooperation techniques are developed to exploit the ICI with reduced backhaul capacity needs. Systems with limited cooperation correspond to two class systems where critical users operate under a cooperative scheme and non-critical users operate under a non-cooperative scheme. The limited cooperation approach has been thoroughly investigated under two autonomous considerations: static and dynamic resource division and user grouping. Various transmission schemes, resource division, resource allocation and user grouping approaches are given offering reduced complexity, tractable and flexible solutions. It is concluded that full cooperation is not required to maximise the overall performance. In addition, with affordable loss in total throughput, the fairness and the performance of the critical users can significantly be enhanced while cooperation is active for only a small portion of total users. Interference-aware resource allocation approaches, that require much less backhaul capacity, are also investigated to mitigate the effect of ICI with much reduced complexity. Two ICIC techniques are proposed. At first, a low complexity heuristic approach is taken in which the base stations seek to limit the ICI power affecting critical users. The second technique assumes a noncooperative power control game approach with pricing in which a penalty is imposed for the ICI caused and this penalty is priced according to the users’ performance metrics. Both techniques have improved the performance of the critical users while securing good performance standards for all other users. In order to place the formerly examined scenarios into an analytical scope, information-theoretic analyses of the per-sector sum rates for multiple linear and nonlinear transmission schemes are established. A modified Wyner model is proposed and a closed form approximation for the multicell joint processing is derived. Moreover, a selective transmission approach is explored to exploit the randomness of time-varying channels.