Research On Cellular Network Resource Allocation Based On Service Characteristics

The popularization of intelligent terminals and the emergency of Internet applications have generated a great demand for wireless access.Mobile data traffic continues growing and energy consumption rapidly increases.An in-depth study of cellular network resource allocation will contribute to making better use of limited wireless resources to meet the increasement of throughput and reduce energy consumption.Describing the service characteristics and understanding the service models have a good reference and inspiration to utilize the cellular network resources effectively.Based on the abundant research of different service characteristics,this paper constructs a multicast transmission model based on user interest and studies an energy-saving scheme with predicted traffic load from the perspective of user interest and traffic data characteristics respectively.By exploiting different service characteristics to allocate the cellular network resource,we aim to increase the system throughput and reduce system energy consumption.First,this paper studies the multicast transmission based on users’interest,introduces the users’ interest matrix,puts forward the concept of social utility,and then considers cellular network resource allocation problem under the constraints of service-user association,subcarrier allocation,power allocation and service transmission rate.At the same time,with the goal of maximizing social utility,the corresponding resource allocation problem is divided into multicast group formation and power allocation.Simulation results show that the proposed scheme can impove the social utility significantly compared with the multicast transmission which does not consider the user interest.Furthermore,this paper considers an energy-saving scheme with predicted traffic load for base station.Based on the existing traffic load information,the characteristics of the cellular network traffic load are analyzed.Then the future traffic load is predicted by using sinusoid superposition model according to the periodic variation.The simulation results show that the sinusoid superposition model can predict the flow well.Next,we propose a grid-based dynamic energy-saving scheme for base station using the predicted traffic load.And an energy-saving scheme which matches to traffic fluctuation is established through selecting some base station into active mode and others into sleep mode at different times.The simulation results with real traffic load show the above model can reduce energy consumption effectively.