| One of the biggest features of Future Wireless Networks(FWN)is multiservice and multi-demand.Unification,intelligence and universe will be the most distinctive manifestation of the feature of FWN.Firstly,various scenarios coexist in the unified network.The network has the ability to satisfy diversified requirements of user simultaneously;Secondly,the demands of communication and computation integrate with each other.The level of network intelligence is continuously improved;Finally,multiple access methods are cooperated in the networks.Globally three-dimensional space-sky-ground wireless networks are everywhere.However,above three characteristics bring great challenges to the ongoing technologies.The characteristics require the network should cooperate the multiple requirements with access methods and wireless resource,as well as consider the influence of user features and system properties on the network capability.The existing wireless network has no ability to solve the resource scheduling problem of multi-service integration under different conditions Therefore,this thesis focuses on the matching problem between multiservice requirements and user resources in FWN.Facing to the FWN characteristics of unification,intelligence and universe,the thesis does the researches on how to realize the coexistence of eMBB and uRLLC,how to realize the integration of communication service and computing service,and how to realize the joint transmission scheduling between satellites and ground BaseStations(BSs)as follows:Firstly,the coexistence of eMBB and uRLLC is an inevitable requirement for the unification of FWN.Different from the eMBB,which requires high data rate,the uRLLC requires very low latency.Two mutually exclusive requirements make it a huge problem how to have two scenarios on the same user equipment(UE)at the same time.In this thesis,a dynamic multiconnectivitybased joint scheduling scheme for eMBB and uRLLC is proposed,where multiple connections can be established between the base station and the UE at the same time,and eMBB and uRLLC are transmitted through different connections.The connections are dynamically adjusted according to the system states and requirements.Multiple base stations can serve a UE at the same time,so that the connection between the base station and the UE and the allocation of wireless resources can meet the service requirements.A modified effective capacity(EC)model for multiconnectivity is also proposed to measure the performance of our framework.By a two-step optimization,the mixed integer nonlinear programming conducted by the EC model is solved.The system-level simulation results show the proposed framework enhances the system throughput effectively while guaranteeing the uRLLC latency.Second,the integration of communication requirements and computation requirements are one of the crucial methods to realize the intelligence of future wireless networks.However,BSs and their computing unit which undertake the functions of resource scheduling and network forecast cannot meet the latency requirement of the computation tasks because of limited computation capability.Offloading the time-sensitive functions to UE is believed as one of the effective methods to tackle this challenge.The disadvantage of this approach is that the computing capability of UE is weak and offloading occupies communication resources.To overcome the above shortcomings,a multi-connectivity enhanced joint scheduling scheme for distributed computation offloading and communication resources allocation is proposed in this thesis.In the scheme,computation tasks are divides into many slices and distributed to UEs to aggregate the computation capability.A communication-incentive mechanism is provided for involving UEs to compensate the loss of UEs,while multi-connectivity is adopted to enhance the system throughput.The offloading failure ratio is also defined as a conclusive constrain for offloading size by analyzing the movement of UEs.Finally,the co-scheduling of offloading size and throughput is solved by a two-step optimization.The system-level simulation results show that the offloading size and throughput of the proposed scheme are larger than comparisons when the time constrain is tight.Third,the combination of ground BSs and low earth orbit satellites is one of the important ways to realize the universe of future wireless networks.However,the different characteristics of base stations and satellites make them show different performance in the face of different services.How to match different services with multiple kinds of access ways will make great influence on the performance of space-ground network.In the thesis,a BS-satellite cooperation scheme using massive MIMO to transmit multiple traffics on multiple connections is proposed to achieve above goal.Firstly,mathematical models are established to illustrate the requirements of high-throughput traffics and high-reliability traffics.High-reliability traffics are model by Stochastic Network Calculus(SNC)and the minimum resources for the latency bound under Rician channel and Rayleigh channel are precisely obtains.The expectation of throughput is taken as the index to measure high-throughput traffics for eliminating the fluctuation caused by small-scale fading.Based on the modeling,the thesis proposes a scheme combining the frequency allocation in every beam and the dynamic multiple connections selection which makes the UE could connect to multiple access point.The scheme could flexibly adjust the frequency splitting according to the network state and service requirements meanwhile optimize the associations between UE and the access point.The system-level simulation results show the strategy satisfies the latency and reliability requirements and improves the system throughput comparing to the comparisons. |
