Theories And Methods Of Resource Allocation Optimization In Cloud-Edge Computing-Based Wireless Networks

To support the extreme performance requirements of low latency and huge capacity in mobile communication networks,cloud computing-based wireless networks and edge computing-based wireless networks are proposed successively.Against this background,the cloud-edge computing-based wireless network takes full advantage of cloud computingbased cooperative transmission and edge computing-based real-time response,which balances the capacity burdens on the fronthaul link and the limited edge computing.However,the characteristics of computing and communication resources are very different in the cloud-edge computing-based wireless network.There is still a lack of theories and methods on how to coordinate the computing and communication resources,which becomes a difficult and hot research topic.Therefore,focusing on fog radio access networks and unmanned aerial vehicle-enabled wireless networks,this dissertation studies the following challenges on the key issue of cloud-edge computing-based differentiated and efficient multi-dimensional resource allocation:(1)The multi-dimensional resources are coupled and it is difficult to characterize the latency:In the cloud-edge computing-based fog radio access network,both cloud and edge nodes possess communication and computation capabilities,in which the networking modes are complex and diverse.The relationship between the transmission and processing latency and the multi-dimensional resources for different modes should be analyzed.(2)The resource optimization is complex and it is difficult to reduce the latency:In the cloud-edge computing-based fog radio access network,the resource allocation should consider optimizing wireless communication and computing resources jointly,which makes the traditional resource optimization algorithms not applicable.The efficient multi-dimensional resource allocation method to minimize the latency should be studied.(3)The network requirements are various and it is difficult to balance different performances:In the cloud-edge computing-based unmanned aerial vehicle-enabled wireless network,the resource allocation is constrained by the energy of unmanned aerial vehicles.The joint trajectory and multi-dimensional resource allocation optimization method should be studied to balance energy consumption and latency.In view of the above challenges,the main contents and contributions of this dissertation are summarized as follows:1.To solve the challenge that the multi-dimensional resources are coupled and it is different to characterize the latency,this dissertation formulates communicationcomputing-cooperative networking modes in the fog radio access network,analyzes the impacts of communication resource,caching content placement,and computing offloading on the transmission and processing latency for different networking modes,proposes a mode selection optimization algorithm.To solve the caching content placement and computing offloading optimization problem,the networking modes are utilized to transform the problem into a multiple-choice multiple dimensional knapsack problem,in which the Lagrangian dual decomposition algorithm is proposed to obtain a local optimal solution,and the branch and bound algorithm is proposed to obtain an optimal computing resource allocation under a specific parameter setting.The simulation results show that the proposed algorithms reveal the impacts of communication and computing resources on the latency,in which the obtained latency is better than that with the greedy computing resource optimization algorithms.2.To solve the challenge that the resource optimization is complex and it is difficult to reduce the latency,this dissertation formulates the networking modes for both uplink and downlink transmissions in the fog radio access network,proposes a joint communication and computing resource allocation optimization algorithm to minimize the transmission and processing latency,analyzes the impacts of multidimensional resources on the mode selection and latency.To solve the latency minimization problem,a greedy algorithm is proposed to obtain the close-form results of bandwidth,caching content placement,and computing offloading.An alternative optimization algorithm is proposed with the networking modes to obtain the mode selection,transmission rate,and processing frequency allocation.The simulation results show that the proposed algorithms can reduce the latency significantly,in which the impacts of communication and computing resources on the mode selection are analyzed.3.To solve the challenge that the network requirements are various and it is difficult to balance different performances,this dissertation proposes a joint trajectory,transmission power control,and processing frequency allocation optimization algorithm in the unmanned aerial vehicle-enabled wireless network,which tackles an energy-constrained latency minimization problem.In particular,the Lyapunov optimization theory is utilized to balance energy consumption and latency,in which an alternative optimization algorithm is proposed to iteratively solve the problem and a low-complexity trajectory optimization method is proposed to minimize the flight energy consumption.The simulation results show that the proposed algorithms have significant advantages in reducing energy consumption and latency,in which a trade-off between energy consumption and latency is revealed.This dissertation researches the theories and methods of resource allocation optimization in cloud-edge computing-based wireless networks.Focusing on the cloud-edge computing-based fog radio access network and unmanned aerial vehicle-enabled wireless network,this dissertation studies multi-dimensional resource allocation methods to optimize the latency,which provides theoretical support for resource management in cloud-edge computing-based wireless networks.