| With the development of mobile communication technologies,smart mobile devices are gaining enormous popularity.By running rich mobile applications on smart devices,users are constantly satisfied with various needs,such as face recognition,natural language processing,augmented reality,virtual reality,and so on.However,complex applications processed on mobile devices will lead to high energy consumption.Due to the small size of mobile devices,hardware resources(eg: CPU computing capacity,storage size,and etc.)and energy supply(battery power)are very limited.Therefore,it is still a challenge to run various computation-intensive applications on devices with limited energy supply and computing resources.Mobile Cloud Computing(MCC)is envisioned as a promising method to address such a challenge.Computing resources of smart mobile devices can be augmented by migrating computation tasks to the resource-rich cloud via wireless access,referred to as computation offloading.Through this approach,mobile devices can offload their tasks to center cloud for computing,which can effectively reduce the energy consumption on mobile devices and prolong the battery life.However,with the increasing number of offloading users,how to selectively choose tasks offloaded to cloud remains an urgent problem to be solved.Therefore,this thesis focuses on how to make energy-efficient strategies in the multiuser offloading scenario.Firstly,we consider maximizing the energy saved among all offloading devices,which is from the perspective of all offloading users;Secondly,we consider maximizing the number of users who can achieve energy saving during offloading,which is from the perspective of each offloading user.Here,the main contributions of this thesis can be summarized as follows:(1)We consider to maximize the energy savings among all devices in multi-user offloading scenario.First,we propose an agent-based MCC framework to enable devices to shorten the average delay of receiving offloading results by filtering process on devices and agents,called AQRF.Moreover,to get offloading strategies among devices,we formulate the problem of Maximizing Energy Savings among All Users(MESAU)under the task completion time and bandwidth constraints.To solve the optimization problem,we propose a Dynamic Programming After Filtering(DPAF)algorithm.In the algorithm,firstly,the original optimization problem is transformed to the classic 0–1 Knapsack problem by the filtering process on the agent.Furthermore,we adopt dynamic programming algorithm to find an optimal offloading strategy.Simulation results show that the proposed framework can get response from agent more quickly than other schemes and the DPAF algorithm outperforms other solutions in energy savings.(2)We consider to maximize the number of beneficial offloading devices in the multi-user offloading scenario.First of all,we adopt OFDM technique and game theory to model each user's pursuit of their own energy-saving purposes,then formulate Multi-User Offloading Game(MUOG)problem.In order to obtain the solution to MUOG problem,Offloading Game Strategy(OGS)is proposed in this thesis,which includes BOT and BOG algorithms.Through the Beneficial Offloading Threshold(BOT)algorithm,we can obtain the threshold that each device can tolerate the number of offloading users.In addition,a group of beneficial offloading users can be obtained from Beneficial Offloading Group(BOG)algorithm.Moreover,it can be proved that BOT and BOG can get an offloading strategy to achieve the Nash equilibrium of MUOG.Through simulations,we compare the model and the algorithm proposed in this thesis with other offloading strategies,and our strategy can benefit more offloading users to achieve energy saving. |
PDF Full Text Request