A Study On Base Station Location Planning Based On Immune Algorithms

In the whole process of wireless network planning, Base Station(BS) location planning is a flexible and crucial problem, is considered to be NP-hard problem. In the case of meeting network coverage and system capacity, how to reduce the cost and improve the profit is the goal of telecom operators. BS location planning is an extremely complex problem, need to be considered the relationship between network coverage, system capacity and construction cost. For the 3G/4G network, the capacity and coverage need to be considered at the same time, this increases the difficulty of solving the problem. In this case, how to select the efficient and reasonable intelligent optimization algorithm to solve BS location problem has aroused wide attentions from scholars at home and abroad.Due to the rapid growth of wireless network users but the growing shortage of spectrum resources, at present, the main bottleneck is the poor coverage, the signal is not stable and so on.Such problems can be solved by increasing the number of BSs, however, BS construction cost is too high, and the complexity is relatively high. Therefore, reasonable BSs construction seems to be particularly important, how to effectively use the limited number of BSs to meet the ever-growing users demand, has became the issues of common concern between science workers and network operators. The model of wireless network BS location planning and corresponding intelligent optimization algorithms were deeply studied in this paper, the main works was as follows:In order to minimize the network construction cost, improve the quality of the user’s service and system coverage, the mathematical model of BS location problem solving was presented, a solution of BS locations based on vector distance immune computation is proposed. The method of antibody concentration calculation based on vector distance was designed, a population initialization program based on opposition-based learning strategy was adopted. The use of improved clone mutation operator, made the clone size can be adjusted dynamically according to the antibody fitness and concentration value, thus well keep the diversity of the antibody population, improve the quality of solutions. Simulation results shows that the proposed methodhas outstanding global and fast convergence ability, can meet the coverage needs and network capacity with low cost of construction relatively.To solve TD-LTE network BS location problem, a novel BS location planning scheme based on combination of immune algorithm and particle swarm algorithm was proposed. First, in the process of problem solving, exploit the advantages of strong evolutionary ability of immune algorithm to make up for the disadvantage of easy to fall into local least value and low convergence precision in particle swarm algorithm. The population diversity of the particles was changed by the immune memory strategy in immune algorithm. Secondly, in mathematical modeling, made the planning area divided into key area planning and general planning area,thereby effectively avoid the waste of resources. Simulation show that the proposed immune particle swarm algorithm has a good ability to overcome the shortcomings of the two optimization algorithms itself, has higher problem solving accuracy and global search ability, can be a good solve to the 4G network BS location planning problem.