Research On Multimedia Resources Allocation Mechanism Based On 5G Networks

The rapid development of 5G mobile communication technology and the popularization of smart mobile terminals allow users to access the Internet anytime and anywhere to obtain desired multimedia services.However,in the construction of informatization,innovative and personalized applications not only increase the user’s activeness in using the Internet,but also increase the user’s ability to drill into multimedia services.Therefore,in order to better meet the needs of users for personalized multimedia services,improve the utilization of multimedia resources,and reduce the network load pressure,this article focuses on the key technology of multimedia resources allocation in the 5G networks as the main research line and carries out the research on multimedia resources allocation mechanism in four aspects:demand forecasting,resources allocation,multimedia service profit and computing task offloading.The main tasks completed are as follows:First,this article selects video users as the research object and carries out research on user demand prediction.Based on Support Vector Machine(SVM)theory,this paper constructs user demand prediction models under different kernel functions,and optimizes the parameters of the SVM video user demand prediction model under Gaussian kernel functions to improve the accuracy of model prediction.Second,in the view of high power efficiency problems faced by the allocation of multimedia resources,this paper comprehensively considers the three factors of system energy consumption,transmission delay and system utility to construct a dynamic multimedia resources allocation model,and proposes a model based on the Lyapunov stochastic optimization theory.The multimedia resources allocation mechanism in the 5G edge cloud networks achieves the optimal resources scheduling through three steps: transmission control,access control and queue update.Finally,the results of simulation experiments verify the effectiveness of the proposed strategy.Third,aiming at the high real-time characteristics of multimedia transmission,this article comprehensively considers system utility,transmission delay and service provider profit issues.From the perspective of service providers,this paper proposes a5 G network-oriented multimedia resources purchase and allocation mechanism.Based on the Lyapunov stochastic optimization theory,the objective optimization problem is transformed into a queue problem.Finally,the results of simulation experiments verify the effectiveness of our proposed strategy.Fourth,aiming at the problem of offloading and distribution of multimedia computing tasks in the 5G networks,this paper constructs a hybrid D2 D cellular network communication model,and proposes a multi-executor cooperative task offloading and computation scheme.The problem is transformed into three independent sub-problems to be solved.Finally,the experimental simulation results verify the good performance of the proposed strategy.