Research Of Power Control Algorithms In Cellular Mobile Communication Systems

Power control is one of the most important approaches to effectively combat thenear-far problem and corner problem, to suppress system interferences and to increasethe system capacity. This dissertation is concerned with power control algorithms inmobile communication systems. The main work of this dissertation is recapitulated asfollows.Firstly, the centralized and distributed power control algorithms are summarized to forma complete theoretic system. Based on the theory and the knowledge of numerical linearalgebra and matrix, power control methods are analyzed in detail. A general frameworkof constructing and analyzing iterative power control algorithms and a general methodof convergence analysis are proposed. Many of the existed distributed power controlalgorithms can be derived as special cases of this basic framework with differentiterative matrices.Secondly, there is a universal problem in the simulation of radio resource management(RRM) algorithms: it needs to construct a channel model and a simulation environmentbeing up to the real radio, which are used to estimate the performance of various RRMalgorithms. In order to solve this problem, this dissertation first analyzes in detail theradio transmission channel and then establishes the computer simulation models ofshadowing and fading by some effective methods. A package of functions is at the sametime developed to create base stations and mobile stations. It becomes convenient tosimulate handoff, rate allocation, power control and other RRM algorithms by suchpackage. This thesis utilizes it in the simulation of power control algorithms and theresults show the effectiveness of the platform.Thirdly, the power control problem is studied from the perspective of energy efficiencyin order to overcome the disadvantageous property. A distributed power controlalgorithm with energy saving property is presented. The energy saving and convergenceproperty are strictly proved by mathematical methods. This algorithm is particularlyappropriate for non-real time data services where the delay tolerance is higher. Thesimulation results also show its energy saving property and convergence. IIIFinally, in digitally power-controlled systems, power levels are discrete. To adapt tosuch limitation, this dissertation put forward a distributed discrete power controlalgorithm based on the second order power control algorithm with better performances.The problem of convergence and uniqueness of this algorithm is well solved. We showthat there is a convergence in a weaker sense, that is, it converges to an envelope ofpower vectors, rather than to a single vector. We also define an optimal power vector towhich the presented distributed discrete second order power control algorithmconverges.