Joint Optimization Of Area Spectrum Efficiency And Energy Efficiency In Ultra-dense Heterogeneous Networks

Ultra-Dense Networks(UDN)can shorten signal transmission distance,improve spectrum efficiency and system capacity by deploying a large number of base stations with lower power and smaller coverage,making it a worldwide recognized green communication technology.The heterogeneous network with multi-type base stations evolves from cellular network with single macro base station,has been widely concerned and applied.Cellular network is usually analyzed in 2D space modeling under UDN technology.As the horizontal densification and vertical densification of the network occurring simultaneously,the height distribution of base stations becomes more diverse,and the network deployment becomes more three-dimensional(3D).As a result,the coverage probability range provided by the traditional 2D model has been inconsistent with the actual network situation.Some studies have compared 3D and 2D models with actual deployed networks and found that 3D models are more accurate.Aimed at the problem that the traditional UDN modeling analysis does not conform to reality and has limitations,this thesis conducts the following research on the network of the dense areas where a large number of low-power base stations are deployed:(1)Based on the 3D Poisson Point Process(PPP)in stochastic geometry,downlink 3-tier ultra-dense heterogeneous network model is constructed.The influence caused by base station density on network performance is analyzed.Firstly,on the basis of this model,the probability density function of the distance between the target user and the base station in 3D space is given.Afterwards,the coverage probability of each layer and the average accessible rate of user in different layers are derived.Area Spectrum Efficiency(ASE)and Energy Efficiency(EE)are derived at the same time,moreover,the impact of each tier base station density on ASE and EE is analyzed.The simulation results show that dense deployment of base stations in UDN can indeed improve SE.However,when the base station density exceeds a certain threshold,ASE growth becomes slow and EE also begins to show a downward trend.The base station density of each layer and the density ratio between layers have great impacts on the system performance.The increase of base station density cannot bring the infinite growth of ASE and EE,so there is a reasonable deployment of base stations to make the overall performance optimal.(2)Considering the trade-off between the key performance of wireless communication network and the problem of optimizing the overall performance of the system,a multi-dimensional efficiency joint optimization problem by combining ASE,EE and Cost Efficiency(CE)is constructed.Subsequently,the objective function combining the three indexes is obtained,and the Firefly Algorithm(FA)of swarm intelligence algorithm is used to optimize the base station density.On the premise that the density of Femtocell is considered as one of the variables,the value of base station density when the system performance is optimal under different conditions is obtained.Therefore,the model established in this thesis is more accurate and effective for the actual deployment,and the research on the influence of base station density on network performance can provide theoretical support for the deployment of base stations.