Study On Interference Alignment In Heterogeneous Networks

In the past decades,mobile communication technology has merging into almost every aspect of people's life and become one the most important infrastructure of modern society.However,traditional cellular networks are no longer capable of providing qualified services due to the rapid growth of data demand from subscribers.Heterogeneous networks(Het Net)which contain high power macro base stations(BSs)and low power small cell BSs are introduced as an effective way to improve spectral efficiency and signal quality,and lower the construction cost.While bringing the aforementioned benefits,the two-tier network structure of Het Net makes the interference more complex and severe which degrade the network capacity.Thus,interference management has become an important subject.Interference alignment(IA),as one of the effective interference management technologies,has also become a hotspot in the study of Het Net.This thesis first investigate the interference condition in downlink Het Net and study the hierarchical IA(HIA)through analysis and simulation.Then the interference problem in uplink Het Net is discussed and a cross-tier IA is proposed and validated through simulations.The mean results and contributions are listed below.1.The interference condition and HIA in downlink Het Net is studied.The design of beamforming matrices of HIA in 2 different access mode of picocell,open subscriber group and closed subscriber group,is derived.HIA adopts a two-stage beamforming design process,and gives a closed form solution under specific scenario which decrease the control signaling overhead in system.The optimized solution of the HIA based sum rate maximization problem is given.Through analysis and simulations,we can conclude that HIA achieves the optimal Degrees of Freedom(Do F),and the sum rate is improved significantly through beamforming optimization.Moreover,the feasibility of Generalized HIA is analyzed.Simulation results also show the appropriate configuration in multiple macrocells scenario.2.The asymmetric interference of macrocell and small cell in uplink Het Net is studied.A cross-tier IA with in uplink Het Net is proposed.Several interfering pair selection(IPS)schemes are given and applied to cross-tier to decrease the load on backhaul links fortransmitting channel state information(CSI).The consumptions of backhaul bandwidth of different IPS schemes are analyzed and compared with no-IPS condition.The feasibility condition of perfect cross-tier IA is investigated.To tackle with the situation that perfect IA is unachievable,semi-definite programming(SDP)is utilized to solve the IA optimization problem.Simulation results show that multi-macrocell cross-tier IA can improve the sum rate of the system effectively.Moreover,simulations on different number of small cell BSs in each macrocell give the appropriate empirical value which achieves best sum rate performance and keeps the load on backhaul at a low level.