Research On Key Technologies Of Resource Management For Next-generation Heterogeneous Networks

With the explosive growth of the traffic demand caused by the rapid evolution of terminals in wireless mobile network,heterogeneous networks(HetNets)have been recognized as an advanced technique to cope with the rapid growth of the traffic generated by billions of mobile devices.Through reasonable deployment of small cells and cooperation between networks,HetNets can push network access closer to customers,and meanwhile satisfy the various quality-of-service requirements of differentiated services.Resource management plays a key role in implementing efficient multiuser communications over complicated wireless network environments,especially the wireless HetNets;thus it becomes an important research topic in the development of wireless mobile network.This dissertation studies the resource management in next-generation wireless HetNets,and the emphases include the reliability and high-efficiency of user communications along with the stability and diversity of acquiring services while deploying and sharing network resources.With the support of the National Natural Foundation of China under Grants61540046,61601347,and 61771366,and the “111”Project of China under Grant B38038,the distinct features of this paper are as follows:1.Research on the issues of communication reliability and service stability for the HetNets resource management.Considering that the influence factors on the reliability of user communications and the stability of acquiring services are different,this study can be divided into two parts.i)First,note that the complicated network environments bring great challenges of providing high-quality services to users.For example,due to the channel fluctuations,limited feedback capacity,and channel estimation and quantisation errors,it is expensive and difficult to estimate the perfect channel state information between a large amount of users and different network access points in HetNets.To further guarantee the reliability of user communications,an energy efficiency resource allocation scheme is proposed based on imperfect channel state information for cognitive HetNets,where both of the outage issue in secondary networks and the collision issue in primary networks are taken into consideration.The scheme jointly optimizes the subcarrier allocation and power allocation to guarantee the quality-of-service requirements for secondary users and primary users in probabilistic manners.Simulation results are provided to show the effectiveness and reliability of the proposed scheme.ii)Second,note that some network access points in the complicated network environments might sometimes inevitably fail,so that the mobile users within the coverage of these access points face service interruption.To keep the stability of acquiring services when abrupt network failure occurs,a cooperative cross-layer resource allocation scheme for self-healing in HetNets is presented to compensate abrupt network failures.The proposed scheme is based on multiple medium access control layer and physical layer of different networks,and is able to ensure the network fairness while keeping the normal communications for customers within the coverage of the failed access points.Simulation results demonstrate the effectiveness and stability of the proposed scheme.2.Research on the issues of service diversity and joint multi-resources optimization for virtualized HetNets.With the increase of the network coverage,the mobile traffics and terminal devices carried by networks become various and abundant.In order to better accommodate the dramatically increasing demand for content delivery and computation,some intermediate nodes within the networks should be endowed with caching and computing capabilities to promote the network performance and the quality-of-experience.i)First,a novel virtualized information-centric HetNets framework is proposed based on wireless network virtualization(WNV)and soft-defined networking(SDN).Through optionally integrating information centric networking function(ICN)or mobile edge computing(MEC)in each virtual network,the purposes of offering diversified businesses and supporting classified services can be realized.ii)Second,according to the designed joint network framework,a caching-assisted MEC mechanism is proposed,where the backhaul bandwidth resource and the computation energy can be conserved through determining whether or not to store the contents before or after the computation.A joint virtual multi-resources allocation strategy is formulated based the proposed mechanism to maximize the overall benefit of mobile virtual network operators.Besides,a distributed algorithm based on alternating direction method of multipliers is adopted to solve the formulated problem with the reduced computational complexity and signaling overhead.Extensive simulations are presented to show the effectiveness of the proposed scheme and the convergency of the presented algorithm.iii)Third,according to the designed joint network framework,a computing-assisted ICN mechanism is proposed,which aims at the issue of the scarce spectrum resource due to the exponentially explosive growth of the adaptive video streaming traffic in future HetNets.Different users can acquire the required video services through storing partial versions of each video content and transcoding to obtain the remaining versions in network,thus enabling the massive content delivery and guaranteeing the delivered content versions to better fit the complicated network conditions and the various mobile devices.Besides,a novel joint virtual multi-resources allocation strategy is formulated based this mechanism,and extensive simulations are also conducted with in-depth analyses.3.Research on the high-efficient communication issue for virtualized HetNets resource management.Based on the previous research,the multicast beamforming technique is introduced to bring possibility of improving spectral efficiency and energy efficiency for the joint architecture thus achieving the high-efficient communications.First,a cache-aware multicast beamforming design is proposed for multicell multigroup multicast,where the multicast mode,the caching strategy,and the network-wide beamforming vector are jointly optimized to minimize the long-term overall energy cost of communications,caching,and computing.Second,due to the high computation complexity of the formulated problem,a computationally-efficient suboptimal algorithm is presented to make the formulated problem tractable,in which the cache-aware multicast approach design and the sparse multicast beamforming design are realized by a step-by-step implementation.Third,in order to promote the stabilization of the system,a robust joint optimization strategy is further designed for the scenario with imperfect channel state information.Finally,extensive simulations are conducted to compare and explain the impacts of different alternative caching methods on the multicast beamforming design scheme.The simulation results demonstrate the effectiveness of the proposed schemes,and meanwhile verify that the presented algorithm is able to significantly reduce the computational complexity with little performance degradation.