Research On Resource Management In CDMA Cellular Systems Based On Successive Interference Cancellation

CDMA cellular systems are interference-limited and one effective way to increase the system capacity is to reduce the multiple access interference (MAI). The receiver in the base station can use multi-user detection (MUD) algorithms to reduce MAI and increase system capacity. Among these algorithms, successive interference cancellation (SIC) is widely employed due to its low implementation complexity and satisfying performance. Although MUD algorithms like SIC can help to reduce MAI and increase system capacity, CDMA cellular system desires more radio resources in response to the rapid increment of mobile users and the high requirements of communication services. This pose a challenge to satisfy the increasing users' requirements based on limited system resources. Hence, in CDMA systems based on SIC, it is also imperative to address the issue of radio resource management (RRM), i.e., subject to Quality of Service (QoS) guarantee, the issue how to fully and efficiently utilize radio resources needs to be carefully studied. In CDMA systems based on SIC, different decoding orders result in different system performance. Consequently, the issue of RRM in this kind of systems is different from those of RRM in systems based on the single user detection (SUD) receiver. In this case, the research of RRM needs to be studied by taking into account the adjustment of decoding orders. This dissertation, from a system design point of view, is devoted to the essential problems of RRM in CDMA systems based on SIC, such as power control, rate allocation, call admission control (CAC) and system capacity analysis under multi-path diversity fading channels. The main contributions of this dissertation can be summarized as follows:We first study power control and related problems in CDMA cellular systems with SIC. The optimal power allocation formula under the multi-cell model is derived. An efficient method for examining the feasibility of systems is proposed, where the transmission (received) powers of users are constrained. The optimal decoding order of minimizing the system outage probability is derived for the case where powers of users are constrained. Finally, the optimal decoding order of minimizing the total transmission power is proposed.In the CDMA cellular system based on SIC, the numbers of possible rate modes and decoding orders grow exponentially with the number of users. How to efficiently allocate rates is an NP problem, which involves the joint optimization of rate allocation and decoding order adjustment. We propose two greedy algorithms for this problem. As a result, both algorithms have a low computation of complexity while achieving good performance. Furthermore, another rate allocation algorithm based on the genetic algorithm is proposed to solve this kind of rate allocation problems. We theoretically prove that the algorithm can completely converge to the global optimum. The simulation results show that the latter algorithm with higher computational complexity can obtain the better solution than the greedy algorithms.CAC is also an important functional module of RRM, next, we turn to study CAC in CDMA cellular systems with SIC and propose two CAC algorithms. One is the CAC algorithm based on upgrading and degrading service levels under the limit of users' transmission (received) power, and the other is the CAC algorithm based on Grade of Service (GoS) to set the admission threshold under both limits of users' transmission (received) power and total received power in base stations. The former helps to maximize the throughput while allowing more users accepted, and the latter benefits to optimize the GoS of system.SIC has heavy impacts on the capacity of CDMA systems, finally, we analyze the capacity of multi-services CDMA cellular systems based on SIC under multi-path fading diversity channels. We take into account the aspects of power control error, shadowing effects, multi-path fading, and multiple diversity channels which are maximal ratio combined by the Rake receiver, etc. The log-normal distributed random variable and Gaussian distributed random variable are used to approximate the total received power in base stations, respectively. The method calculating system capacity in this situation is obtained. Unlike the methods in existing literatures, we avoid the simplified or unrealistic assumptions during analysis such that the corresponding results can be used as a reference for system capacity planning and design.