Research On Distributed Resource Allocation And User Access In Ultra-dense Networks

With the rapid development of mobile communication and Internet tech-nologies,a large number of end devices such as smartphones,tablets,and wear-able devices have increased at an unprecedented speed.These end devices are eager for high-speed and low-latency access services.At the same time,as the popularity of mobile applications surge,mobile network traffic is exploding.In order to meet the demand for massive-connection and high-speed,ultra-dense network,which deploys small-cell base stations on a large scale to achieve spa-tial multiplexing gain,becomes a fundamental enabling technology for the fifth generation mobile networks(5G).In order to give full play to the advantages of ultra-dense networks,there are still some key problems that need to be addressed,including access selec-tion caused by the increasing number and type of small-cell base stations around users,interference caused by overlapping coverage among small-cell base sta-tions,and computing task offloading problems caused by the rise of computing-intensive applications,etc.Resource allocation and user access technologies can effectively improve system performance and user's service experience by properly scheduling resources and flexibly controlling user access.Besides,due to the large-scale deployment of small-cell base stations,the operation com-plexity and signaling overhead of centralized management scheme are dramat-ically increasing,and it is necessary to investigate the management solutions suitable for ultra-dense networks.Therefore,we focus on resource allocation and user access technology in ultra-dense networks,with joint consideration of non-orthogonal multiple access(NOMA),self-optimization,and edge comput-ing technologies.We propose efficient distributed resource allocation and user access schemes to address the access selection,interference,and computing offloading problems,respectively.These proposed schemes could effectively improve system capacity,spectrum efficiency,and energy efficiency,reduce interference,and meet user quality of service(QoS)requirements.The main contributions and innovations of this paper are as follows:1.A user-centric resource allocation and user access scheme is proposed.Firstly,we introduce the NOMA technology into ultra-dense networks and dynamically group multiple access points(APs)into an AP group to support one user on the same spectrum resource based on user requirements and dy-namic network environments.In this way,system throughput and spectrum efficiency can be greatly increased.Secondly,we formulate the resource allo-cation and user access selection problem as a non-convex combinatorial opti-mization problem with joint consideration of the access and backhaul links.For tractability purposes,we decompose the problem into two sub-problems and propose low-complexity algorithms based on matching and disciplined contour programming(DCP)theories to obtain a feasible solution.Finally,simulation results show that the integration of NOMA and ultra-dense networks can effec-tively improve spectrum efficiency.Also,simulation results demonstrate that the proposed joint resource allocation and user access scheme can effectively reduce outage probability and improve system throughput.2.An interference-aware distributed resource allocation and user access strategy is proposed.Firstly,we introduce the self-optimizing technology into ultra-dense net-works.Considering the dynamic and complicated interference conditions,each small-cell base station with cognitive abilities constructs a local interference graph and adaptively performs user access selection and resource allocation based on the requirements of users.Secondly,we formulate the interference-aware resource allocation and user access problem into a mixed integer nonlin-ear programming problem.For tractability purposes,we decompose the prob-lem into two sub-problems and propose distributed algorithms based on interfer-ence graph coloring and non-cooperative satisfaction game.Finally,simulation results show that the integration of self-optimizing and ultra-dense networks can effectively reduce outage probability.Also,simulation results demonstrate that the proposed distributed resource allocation and user access strategy can effectively reduce the interference and improve the system throughput.3.A distributed delay-limited resource allocation and user application of-floading mechanism is proposed.Firstly,we introduce the edge computing technology into ultra-dense net-works.Considering the requirements of computing-intensive applications,each fog node jointly performs user access selection,offloading decision,and re-source allocation for providing offloading services for users.Secondly,consid-ering the collaborative offloading of computing tasks between fog nodes and cloud servers,we formulate joint resource allocation,user access selection,and computing task offloading problem as a mixed integer non-convex optimization problem.For tractability purposes,the original problem is transformed into a convex optimization problem by variable relaxation and variable replacement.Then,we propose alternating direction method of multipliers(ADMM)based algorithms to solve the problem in an efficient and distributed way.Finally,simulation results show that the integration of edge computing and ultra-dense networks can effectively reduce user delay.Also,simulation results demon-strate the proposed resource allocation and user application offloading mecha-nism can effectively reduce energy consumption and improve resource utiliza-tion.