Research On Resource Allocation Algorithms For Mobile Edge Computing

In recent years,with the rapid development of mobile communications and the increasing popularity of smart mobile devices,mobile communication services have increased dramatically,However,mobile terminals are resource-constrained devices,traditional cloud computing methods have been unable to meet the increasingly rich business needs.In order to improve the quality of experience(QoE)of user,Mobile Edge Computing(MEC)becomes one of the key technologies in the fifth generation of mobile communication(The Fifth Generation,5G)system,providing an effective solution in the reducing the pressure of core network and the improvement of user experience,so it has been widely concerned by the academic and industry community.Focusing on the resource allocation in MEC for research,this thesis proposes a joint allocation algorithm for computing communication resources based on pricing mechanism and the strategy of video buffer update and allocation based on adaptive bit stream.The main work of this thesis is as follows:(1)A computing communication resource replacement algorithm based on pricing mechanism is proposed for MEC,which considers user service quality comprehensively and communication network status,and provides multiple computational offload paths.This algorithm establishes a resource allocation model in a multi-cell scenario using the limited computing and communication resources of the MEC server.Based on QoS guaranteed resource allocation algorithm and effective pricing strategies,it provides users with adjustable traffic service and selects the optimal computing offload path to provide services for the user,so that the macro base station and each LTE small base station are load balanced,as well as maximized network revenue of MEC operator.The simulation results show that the algorithm can effectively improve the resource utilization and user experience compared with the traditional resource allocation algorithm.(2)A video buffer update algorithm based on adaptive bit stream in MEC system is put forward.The distributed storage technology is used to effectively manage massive data.The video stream is distributed on the HDFS storage nodes of each MEC server,which facilitates the classification and extraction of data information by using artificial intelligence and other algorithms.By obtaining the request resources from the local and neighboring MEC caches and using the efficient computing power of the MEC server,Based on different communication costs on contents from MEC servers,this algorithm would select the optional storage node,An adaptive bit rate(ABR)-based algorithm is proposed to transcode and transmit between different bit rates based on cache conditions and user request content.Simulation results shows that the access delay and video return rate are reduced in the premise of ensuring the cache hit ratio compared with the existing algorithms,this algorithm improves the utilization of MEC storage resources.(3)A simple MEC verification system is put forward,which consisted of a video server and a mobile terminal and verifying the MEC video transmission.The test results show that the MEC system greatly reduces the video transmission delay compared with the traditional cloud platform.