Research On Task Allocation Strategy For Mobile Computation Migration On Cloud Base Stations

The traditional centralized cloud has the advantage of strong computing power,convenient management and great scalability.However,the latency gets large and unstable when the distance between mobile terminals and cloud is far away,which restricts the development of mobile computation migration largely.Existing task allocation strategy is mostly aimed at single end users,without fully considering the different sensitivity and will of cost to terminal energy consumption between different users.The thesis combined them together and enabled traditional mobile base stations with computation and storage abilities.The thesis also proposed task assignment strategy from both users and operators'perspective and proved its effectiveness by simulation.The research contents and main results are as follows:(1)The thesis studied on the disadvantages of traditional centralized cloud and the feature of mobile radio communications,enabled traditional mobile base stations with computation and storage abilities.Compared to the traditional centralized cloud,the cloud base station is nearer to mobile terminals and thus improves the migration latency and stability.(2)The abilities of communication,storage and computing vary from one cloud base station to another,while different terminal tasks have various requests for computing,communication abilities and latency restriction,making it important to allocate resources in a rational manner.This thesis proposed a task allocation strategy based on cloud base stations.The strategy took the utility functions of different users into account and realized the rational distribution between computing resources and huge computing tasks,maximizing the benefits of all mobile terminals.The simulation proved that the overall interests are significantly higher than the traditional centralized allocation strategy.(3)The implement of cloud base stations needs mobile operators additional investment and the cost for migration tasks varies between mobile terminals.Therefore,how to get the biggest interests with the smallest investment on the basis of meeting the users' restrictions is the core problem.The thesis proposed a task allocation strategy based on cloud base stations environment which can maximize the mobile operators' benefits by using utility functions of different users.Simulation results showed that the interests of operator are magnificently higher than the traditional centralized tasks allocation strategy without using utility optimization.The thesis focused on the task assignment problem,proposed two task assignment strategies from both users and operators' perspective and proved its effectiveness.