Modified Particle Swarm Optimization-based Network Optimization

With the wide application of 4G network in China,4G base station construction speed stride forward,The network optimization of FDD-LTE also needs to be carried out.4G network in the current situation is not officially supported voice function,Network speed is the only benchmark for measuring the network situation.Peak data up to FDD-LTE of 150 Mbps compared with the peak data is only TD-LTE of 100 Mbps,Regardless of the network speed,Or at home and abroad to the extent of the application,Have more feasibility study?However the system capacity is limited,and in block of flats,mountainous areas and other special regional environment,to ensure the coverage and speed performance of the 4G network is a problem worthy of further research.In this paper,the main factors that affect the network quality are obtained through the methods of data collection,traffic analysis,hardware checking and so on.Comprehensive consideration of multiple aspects of network parameters,such as coverage requirements,network capacity,adjustment costs,Quality of Service(QoS),etc.,to make a tradeoff between multiple objectives.Modeling the network optimization problem as a mathematical problem.Due to the solution space of network optimization problems may be large,and involve many network parameters,hand to use closed form expressions to make an accurate description.It constitutes a NP-hard problem,better to use intelligent search algorithm for its evolution.This method can optimize the network quickly and accurately and efficiently.Specificly optimize the network by adjusting the antenna tilt angle(ATA)of evolved Node B(e NB).The evaluation indexes of coverage optimization is the number of mobile stations(MSs)served by eNBs,which is measured by the reference signal that mobile station(MS)obtained,and optimize the coverage problem by maximizing the number of MSs served.In the improved particle swarm optimization algorithm,there is a group of a collection of particles can be regarded as antenna angles,fitness function is defined as the total number of MSs served,and each iteration of the evolution speed corresponds to the adjustment of antenna angle scale.By appropriate modification of the above parameters or equipment,make little overhead as far as possible with the optimal network quality at the same time,so that the network has made full and effective utilization of resources.Based on theoretical analysis and practical case verification,this paper proposes an improved PSO algorithm to adjust the parameters of the antenna,which can optimize the performance of FDD-LTE network effectively.