Graph-based Interference Management Technology In Heterogeous Networks

In recent years,the world-wide mobile traffic has shown explosive growth.Therefore,developing new technologies to improve the wireless network capacity is one of the most urgent demands for operators.To cope with such challenge,the method that establishes heterogenous networks via deploying low-cost low-power small cells in hot spots has been proposed.The deployment of small cells can effectively fill the coverage hole of the macro cell and improve the spectrum reuse,so that the transmission rate in hot spots can be significantly enhanced.The unplanned use of heterogenous networks,nevertheless,would cause serious co-tier and cross-tier interference,which severely hinders the improvement of network capacity.Hence,to study effective interference management methods becomes an essential matter.It is known that interference management for large-scale networks is generally of extremely high computational complexity.The graph-based interference management strategies can always reduce the complexity efficiently,and therefore they are widely used.In this dissertation,we conduct graph-based studies on three key challenges of the interference management strategies in heterogeneous networks,shown as follows:1.In view of the contradiction between the inflexible interference management and the dynamic user requirements,with considering the cross-tier and co-tier interference in the heterogenous networks,we investigate the joint ORA and interference coordination scheme based on dual time-scales.Specifically,the resource allocation scheme is adjusted dynamically under two time-scales: During the large time-scale,the RBs are roughly assigned to every single cell with centralized algorithm;during the small time-scale,the RBs owned by each cell are allocated to its attached users in a sophisticated way.For maximizing the minimum of users' throughput satisfaction rate(TSR),we formulate optimization problems in two time-scales,followed by Exhaustive Search based Resource Allocation(ESRA)algorithm and Distributed Resource Allocation(DRA)algorithm to solve them respectively.Simulation results manifest the superiority of our proposed scheme over inflexible interference management in guaranteeing the fairness of users' TSR.2.In view of the contradiction between the fairness of interference management and the unfairness of user priority,with considering the co-tier interference in the heterogenous networks,we investigate the admission control and resource allocation schemes based on user priority.Specifically,the users in the heterogenous network are divided into two types: High priority(HP)and low priority(LP),of which HP ones have higher precedence to access the network and enjoy high-resolution video streams.Then in order to eliminate the co-tier interference with maximizing the number of quality of service(Qo S)guaranteed users,we formulate the joint admission control and resource allocation problem into an integer nonlinear programming(INLP)problem,followed by a graph-based algorithm with polynomial computational complexity to solve it.The basic principle behind the proposed algorithm is to firstly chordalize the given conflict graph,and then perform admission control with Priority Differentiation based Admission Control(PDAC)sub-algorithm,followed by the final Resource Block(RB)assignment with Maximum Effect Rank Allocation(MERA)subalgorithm.Simulation results show that the proposed algorithm can achieve the maximum number of Qo S-guaranteed users and ensure the privilege of HP users.3.In view of the contradiction between the traditional interference management and the new-style network architecture,with considering the intra-cell co-tier interference in the heterogenous networks,we investigate the joint user matching and resource allocation scheme based on the full-duplex enabled small base station(SBS).Specifically,suppose that the SBS is equipped with full-duplex radio so that it can simultaneously serve a pair of uplink and downlink users on the same frequency.An optimization problem is formulated with the goal of maximizing the number of satisfied user pairs,of which multi-users' Qo S requirements are taken into account.Then a multi-step matching algorithm is presented,which establishes the feasible graph according to the power restrictions,and employs the Hungarian method to search the maximum matching for the feasible graph.Finally,simulation results verify that our proposed algorithm achieves the maximum number of satisfied user pairs with polynomial complexity.