The Research Of The Green Wireless Cellular Network Planning

With the rapid development of wireless network, wireless network planning and optimization in which base station location is the key link can be a long-term and arduous challenge which will be faced with various telecom operators. With the increasing number of mobile users, types of business continues to increase, the quality of service requirements continue to increase, the network layout planning requirements are also increasing. And among them, the quality of site planning, optimizing directly determines the quality and performance of the wireless network. Therefore, we should pay more attention to the importance of base station planning and optimizing in wireless network.After studying the knowledge of wireless network planning, we resolve the coverage problem of wireless network in this thesis. Firstly, we build simulation model of wireless network based on multiple type of base stations, and solve the problem by Gurobi, which is a linear solver. In this process, LP which is a type of file format is used. At the same time, we propose a new algorithm to solve the problem, and to verify the effective of the new algorithm. We use a heuristic algorithm-GA(genetic algorithm) to solve this problem too. After comparing the above three kinds of solving methods, we draw the conclusion that compared with the genetic algorithm, the new algorithm has some improvement in solving performance;and compared with the Gurobi, although there is a gap between results, there is a big advantage in running time. Due to the obvious shortcomings in the visualization of the results of above three kinds of solving methods, we use AIMMS to solve the model again. This part will be introduced following the steps of operation for AIMMS which is a novel optimization software. Firstly, we will construct a model tree which is used to represent the network model, the most important parts of model tree are:S(set). P(parameter)、V(variable) and C(constraint). Then we solve the model tree after building it, After solving the model, we can get a interface for the display of the results of the solution by adding elements of the model tree. So, we can cognize the results clearly and conveniently.Finally, a summary is given, and the shortcomings of this paper and further research directions were proposed, the experimental result should provide a guideline in the future network deployment for the operators in the effort of BS placement and cost evaluation.