Research On Intelligent Scheduling Technologies Based On Optimizing The Energy Efficiency For Heterogeneous Converged Network

With the popularity of mobile Internet, the network convergence technology has drawn more and more attention as an effective way to overcome the contradiction between the rapid growth of mobile data traffic and the limited transmission capacity of wireless network. Furthermore, from the perspective of green communications, the energy efficiency(EE) has become one important performance metric of wireless networks. Aiming at intelligent scheduling for multimedia contents and focusing on the EE of the broadcast-cellular heterogeneous converged network(HCN), stochastic geometry theory and Poisson point process are applied to model and analyze the position of base stations and users for the wireless communication network in this paper.In order to achieve higher network throughput and EE, and make the best use of the UHF idle spectrum resources after the transformation of existing digital broadcast network, the HCN which consists of micro-cellular network and digital broadcast network provides an effective way to meet the rapid growth for future wireless mobile broadband network. In the converged network, we define the total number of multimedia contents and the proportion(or the number) of the multimedia contents dispatched to the broadcast network by the content scheduling center as M and per(or n), respectively. The basic scheduling strategy for the multimedia contents requested by the users is that according to the request probability which is sorted descending for multimedia contents, the greater request probability of ê?M* per ú? or n kinds of multimedia contents are transmitted by the broadcast network and the remaining are transmitted by the cellular network. Research on converged networks is divided into the following two aspects:1. From the perspective of the user requesting multimedia contents, the contents requested are divided into public contents and personalized contents according to the probability, which is sorted descending in Zipf distribution, of multimedia contents requested by users. Based on random geometric space network model, we consider the coverage areas of the broadcast network, the active probability of cellular network base stations and the outage probability of users, derive the effective capacity of broadcast-cellular HCN, and calculate the EE of converged networks. Using per as a reference variable, this paper analyze its impact on the EE of the converged network. Simulation results show that when the scheduling proportion per is fixed, the density of microcell base station deployed in converged network is less than user density so that the EE of converged network is higher than that of cellular network under the same parameter configuration.2. Since the expression of the EE of converged network in chapter 3 is too complex, it is difficult to give the optimal value of per from the viewpoint of theoretical analysis. To this end, we try to simplify the EE expression. In the simplified scenario for converged network, aiming to optimize the energy consumption, we take the number of multimedia contents transmitted by broadcast network and bandwidth allocation factor as constraint, and introduce Lambert W function to solve this optimization problem. It is shown that with the increase of the total number of multimedia contents, the simulation results of the number of multimedia contents transmitted by broadcast network and the factor of bandwidth allocation match well with the theoretical ones, which proves the correctness of solutions to the optimization problem.