Research On Optimization Of Tracking Areas In LTE Networks

Optimization of LTE tracking area refers to the improvement of the current network tracking area allocation scheme,which improves the network performance indicators,such as the number of handoff among tracking areas,the paging volume and the paging success rate.In recent years,the optimization of LTE network tracking area has just started,and the research on optimization of tracking area needs to be perfected.In this thesis,Multi-Objective Evolutionary Algorithm based on Decomposition(MOEA/D)and differential evolution are used to solve the problem of tracking area optimization in the LTE network.The main work is as follows.(1)A formal model of multi-objective and multi-constraint LTE tracking area optimization is established.By analyzing the planning and optimization principles and the distribution conditions that should be followed in the process of tracking area design,this thesis proposes three goals of tracking area optimization,including minimizing the number of handoff among tracking areas,minimizing the paging volume,and maximizing the paging success rate in network.(2)MOEA/D is used to solve the tracking area optimization model.Besides,the evolution operators such as crossover,mutation and selection are improved by differential evolution.After the evolution is completed,the elite solutions are filtered by Pareto ranking and preference-based selection to select the best individuals as the optimization result.The improved strategies,evolutionary steps and the results of two algorithms,i.e the genetic algorithm and differential evolution are analyzed in detail.Evaluation results show that both MOEA/D and differential evolution can effectively optimize the tracking areas to which the optimized cells belong.Due to the randomness of the genetic algorithm and the convergence speed of the differential evolution algorithm,the result of the differential evolution is better than that of the genetic algorithm in the case of multiple scale comparison.(3)A clustering-based network decomposition algorithm is proposed.Before optimizing the tracking area of the optimized cells,the algorithm preprocesses the cells to be optimized,which improves the effect of tracking area optimization.According to the classification of the handoff relationship of cells,it is expected that the handoff amount between cells within the same cluster is larger,and the handoff amount between cells within the different clusters is smaller.The cells within the same cluster should belong to the same tracking area.In order to ensure that the cells in the same tracking area are geographically directly adjacent to or indirectly connected to each other,the Voronoi diagram is used to construct the geographic neighbor relation among cells.Evaluation results show that the process of optimization of the tracking area has ensured the connectivity of the district.(4)The LTE network tracking area optimization software was designed and developed.The data of the actual tracking areas of various scales was used to verify and analyze the correctness and effectiveness of MOEA/D and differential evolution.The evaluation results show that the MOEA/D-based tracking area optimization method effectively improves the performance of the actual network and achieves good results.As the above-mentioned research,the optimization method based on MOEA/D can satisfy the requirement of tracking areas optimization in LTE networks and it provides a high quality solution to optimize the tracking areas and has a wide application prospect of the research on LTE network tracking areas.