Research On Resource Management In 5G Heterogeneous Networks

With the rapid development of wireless communication networks,mobile data has been explosively increasing,and various applications have been e-merging.5G heterogeneous networks emerges at this historic moment.From the perspective of service applications,5G networks need to meet the hetero-geneous user requirements with large traffic,ultra-dense connections,and fast mobility.From the perspective of communication technology,the introductions of mobile edge caching,spectrum sharing,device-to-device(D2D)communi-cation,ultra dense networks(UDN)technology,etc.,have led to the coexis-tence of multiple heterogeneous technologies.From the perspective of spec-trum bands,there are multiple heterogeneous spectrums such as authorization bands and non-authorization bands in 5G networks.The existences of these heterogeneous service requirements and differentiated technologies bring new opportunities for resource management in 5G heterogeneous networks,and also pose great challenges for the improvement of network performance and guarantee of user experience.Therefore,it is of great significance to study the resource allocation in 5G heterogeneous networks.The innovations and contributions of this thesis can be divided into the following sections:1.Resource management for heterogeneous user requirements.In ultra-dense networks,a two-stage channel and power allocation algo-rithm is proposed to achieve different performance goals of cluster heads and cluster members in ultra-dense networks for the balance of spectrum efficiency and energy efficiency under the same kind of quality of service(QoS).Cluster-ing based on mutual interference of small cells is proposed to reduce interfer-ence between small cells.Different from the existing clustering resource allo-cation algorithms,considering the different performance requirements of clus-ter heads and members,and their non-cooperative competition relationship,the resource allocation problem is modeled as a multi-leader multi-follower Stack-elberg game.Through simulation analysis,it is verified that the the two-stage resource allocation algorithm greatly improves the system energy efficiency under the condition that the system throughput decreases within an acceptable range.Since there are many emerging services in 5G networks,which have their own differentiated requirements.Therefore,ensuring different service needs is an important issue in resource management.Based on such considerations,in D2D-assisted cellular networks,a hierarchical power allocation algorithm is proposed for heterogeneous statistical user requirements.In order to accurate-ly represent the difference in service requirements,cellular users with delay-tolerant services and D2D users with delay-sensitive services are separately analyzed,and energy efficiency models are established with the concepts of ef-fective capacity and outage capacity.Based on the non-cooperative relationship between cellular users and D2D users,and the characteristics of time-varying Rayleigh fading channels,a power allocation algorithm based on hierarchical game is proposed.Compared with the existing researches,we consider the im-pact of different delays and outage requirements on resource allocation in the time-varying channel environment.Through the simulation results,the hierar-chical power allocation algorithm can significantly improve the system energy efficiency while satisfying the differentiated requirements of users.2.Resource allocation based on heterogeneous technologies.While multiple technologies in heterogeneous networks improve network performance,they also pose great challenges for resource allocation due to interactions between them.Different from the existing researches,in D2D-assisted cellular networks,the hybrid cache transmission algorithm and the joint cache transmission algorithm are proposed respectively to improve the system energy efficiency by using different cooperation modes of base station caching and D2D caching in the mobile edge caching.In D2D-assisted het-erogeneous networks,the social characteristic of D2D communication is used to realize the cooperation of mobile edge caching and D2D communication.Through the joint transmission of D2D caching and base station caching,a user-oriented resource allocation algorithm is proposed.3.Resource allocation with heterogeneous spectrums.In heterogeneous spectrum networks,this thesis proposes a low complexi-ty distributed channel allocation algorithm.Considering the dense overlapping deployment of small cells,this thesis first constructs an inter-cell interference graph through graph theory.Then,the distributed features of the potential game are used to complete the heterogeneous spectrum cooperation between the dense small cells,while the service experiences of small cell users are guar-anteed.Service satisfaction does not increase significantly when the user rate exceeds a certain threshold.Based on this theory,this thesis uses the coalition-al game to realize the channel sharing between D2D users and small cell users to improve the spectrum utilization.Through theoretical analysis and simu-lation verification,the proposed distributed algorithm can significantly reduce the complexity of resource allocation and greatly improve the user experience and system throughput.Finally,the research work and future research directions are summarized.