On Uplink Base Station Cooperation In Mobile Communication

In current mobile communication, base station (BS) cooperation, also referred to as coordinated multi-point processing or transmission (CoMP), has emerged as a key technique to combat with inter-cell interference and improve the network throughput and user fairness.This thesis investigates several key techniques in uplink BS cooperation, covering many basic problems including sub-carrier resource allocation, power control, backhaul capacity allocation, cooperative strategy selection etc. Some practical restrictions, such as backhaul capacity limitation, are also considered. Apart from theoretical analysis, low-complexity algorithms are designed and validated through simulations.Firstly, the scheduling problem in uplink BS cooperation is studied. The sub-carrier allocation problem and the power control problem in uplink coordinated schedul-ing problem are jointly considered. A low-complexity algorithm is proposed, which is able to obtain a sub-optimal solution with linear time complexity. Moreover, three con-cepts, namely neighboring BS set, user filter and BS filter, are introduced to further reduce the complexity and accelerate the convergence of the proposed algorithm.Secondly, the problem of combining different cooperative strategies in uplink BS cooperation is studied. We consider adopting several decentralized strategies in a coop-eration set to fully utilize the asymmetric channel quality and the backhaul infrastructure and maximize the network performance. A heuristic algorithm with polynomial time complexity is proposed to establish the combination of the multiple cooperative strate-gies, which consists of two procedures, namely strategy selection and backhaul capacity allocation.Finally, the influence of backhaul capacity limitation on cooperative interference cancellation is analysed. The error propagation procedure is modeled with a Markov chain, with which the composition and probability distribution of the residual inter-ference under different cooperative interference cancellation techniques are analysed. Closed-form expressions for the symbol error probability are derived and validated it with link-level simulations.