Performance Analysis For 5G User-centric Network Via Stochastic Geometry

5G User-centric network (UCN) has shifted the traditional network architecture from cell-centric to user-centric. UCN organizes user-specific dynamic transmission point group (TPG) to provide service for users and eliminate cell borders. By reforming the TPG, the network will guarantee the consistency and continuity of service. Using the tools of stochastic geometry, this thesis analyses the throughput of small cell with non-ideal backhaul in ultra-dense network, the outage probability and average rate of UCN and outage probability for line-of-sight (LoS) and non-line-of-sight (NLoS) transmission in UCN, which will provide guidance for the deployment of UCN. This thesis mainly contains three topics:For the dense deployment of small cells, a semi closed-form expression, which analyses the throughput of small cells with limited backhaul capacity, is derived. To approximate the analysis, a backhaul-aware biasing adjustment model is proposed for traffic offloading and flexible coverage.For the user-centric network with dense deployment of transmission points, this part firstly analyses the outage probability and average achievable rate under general assumptions and derives simple closed-form expressions for Rayleigh fading in UCN, which gives a theoretical performance boundary. The comparison with the performance of ultra-dense network demonstrates the improvement in UCN.Numerical simulations are performed to verify the accuracy of the analytical results and to provide guidance on the size of TPG.For the hotspot deployment in UCN, the outage probability of UCN is analyzed with the consideration of a sophisticated path loss model of 3GPP, which incorporates both LoS and NLoS transmissions. The simulation results verify the accuracy of analytical results and demonstrate the impact of transmission point density and the transmission of LoS and NLoS on network performance.