Research On Dynamic Resource Allocation Strategy Based On C-RAN Architecture

As one of the key technologies of 5G cellular systems,the Cloud Radio Access Network(C-RAN)architecture can not only support growing user demand,but also cut down capital and operating expenses of network operators.In addition,efficient resource management decisions under the C-RAN architecture can further improve energy efficiency,revenue of the service provider,and Quality of Experience(Qo E)of the user.Therefore,this thesis focuses on the dynamic resource allocation problem based on the C-RAN architecture.The main research contents and innovative works are summarized as follows:1.Aiming at the optimization of resource management methods under the network architecture of hybrid energy supply based 5G heterogeneous cloud wireless access network(H-CRAN),this thesis proposes a dynamic network resource allocation and energy management scheme based on deep reinforcement learning.Firstly,this algorithm takes into account the volatility of the arrival of renewable energy and the randomness of the user's data services.At the same time,in order to satisfy the stability of the system,the sustainability of energy and the Quality of service(Qo S)requirements of users,a Constrained Markov Decision Process(CMDP)model is established to maximize net revenue of system service provider.Secondly,the proposed CMDP model is transferred into a unconstrained Markov Decision Process(MDP)problem based on the Lagrange multiplier method.Then,considering that both the action space and the state space are continuous value sets,based on Deep Deterministic Strategy Gradient(DDPG)algorithm,the algorithm jointly optimizes resource allocation and energy management strategies.Finally,simulation results show that the algorithm can improve net revenue of the service providers and system energy efficiency while ensuring network stability,sustainability of energy and Qo S requirements of users.2.To optimize strategy of resource allocation and task offloading decision on D2 Dassisted cloud-fog architecture,a joint resource allocation and offloading decision algorithm based on a multi-agent architecture deep reinforcement learning method is proposed.Firstly,considering incentive constraints,energy constraints,and network resource constraints,the algorithm jointly optimizes wireless resource allocation,computing resource allocation,and offloading decisions.Further,the algorithm establishes a stochastic optimization model that maximizes the total user Qo E of the system,and transfers it into an MDP problem.Secondly,the algorithm factorizes the original MDP problem and models a Markov game.Then,a centralized training and distributed execution mechanism based on the Actor-Critic(AC)algorithm is proposed.In the centralized training process,multi-agents obtains the global information through cooperation to optimize the resource allocation and task offloading decision strategies.After the training process,each agent independently allocates resource allocation and task offloading based on the current system state and strategy.Finally,the simulation results demonstrate that the algorithm can effectively improve user Qo E,and reduce delay and energy consumption.