Research On Data Offloading And Energy-aware Optimization Technology In Mobile Edge Computing

With the development of wireless networks and the vehicular networks,the new generation of smart mobile terminals will generate massive amounts of data and require multiple service support.Therefore,the security and Quality of Service(Qo S)requirements for computation and communication are more stringent,and low latency and high reliability are essential.However,the computation capability of mobile terminals cannot meet the requirements of computation and power consumption.In ultra-dense computation and communication environment,it is difficult to ensure the real-time reliability of task completion.Although the battery capacity and computing power of mobile terminals have been greatly improved due to advances in processor technology and manufacturing standards,considering a huge mass of computationalintensive or latency-sensitive workloads,their processing ability is insufficient to meet the Qo S,such as delay requirements,energy consumption requirements.Due to the exponential increase in the amount of computation data generated by the mobile terminal under stringent deadlines,the battery capacity and resource constraints remain a bottleneck.Understanding how to meet current computation requirement for addressing this problem is a key challenge.In previous years,cloud computing,which offers tremendous computing resources,was able to meet the computation demand of wireless terminals and improve user experience.There has been some research on computation task offloading,wireless resource allocation in cloud computing.However,the intractable problem in communicating with mobile cloud servers is the backhaul delay of downlink transmission,which causes delay-sensitive tasks unable to meet the delay requirements.The emerging mobile edge computing(MEC)has recently become a computation and communication technology commonly used in wireless networks or vehicular networks,because MEC servers are physically located closer to mobile terminal devices and can provide supercomputing capabilities.Therefore,if the computation tasks generated by the mobile terminal in real time are offloaded to the MEC server,and the server returns the processing result to the mobile terminal after receiving the offloaded tasks,this solution can usually meet the latency constraint.However,there are still some new problems that need to be further studied in the offloading of MEC,such as: base station selection in time-varying channel states,separable task processing,and energy awareness.These problems are still challenging to a great extent.This study conducts further research on offloading in MEC.First,the study optimizes the offload selection problem of multiple edge servers under time-varying channel conditions in heterogeneous wireless networks,and then further performs partial offloading for the separable computation tasks generated in vehicular networks environment to minimize the long-term cost.Finally,the joint optimization of computation offloading and energy aware in MEC is considered.The main research of this study is as follows:Firstly,owing to their limited computing power and battery level,wireless mobile terminals can hardly handle computation-intensive workloads by the local processor.Depending on MEC servers with sufficient computing power and communication resources connected to base stations,the study designs a framework with multiple static and vehicle-assisted MEC servers to handle the workloads offloaded by mobile terminals.For obtaining the optimal computation offloading scheme to minimize the weighted long-term cost,the study models the offloading decision optimization problem as a Markov decision process(MDP).Then,a partial computation offloading scheme based on reinforcement learning(RL)is proposed to address the absence of priori knowledge.The proposed scheme can learn the optimal offloading decision based on stochastic workload arrival,the changing channel state,and the dynamic distance between mobile terminals and the edge servers.Moreover,to avoid the curse of dimensionality caused by the complex state and action spaces,this study presents an improved computation offloading method based on deep reinforcement learning(DRL)to learn the optimal offloading policy using deep neural networks.Extensive numerical results illustrate that the proposed algorithms based on RL and DRL can autonomously learn the optimal computation offloading policy with no priori knowledge.Secondly,the vehicular network needs efficient and reliable data communication technology to maintain low latency.It is very challenging to minimize the energy consumption and data communication delay while the vehicle is moving and wireless channels and bandwidth are time-varying.The vehicles and roadside units(RSUs)can offload computing tasks to MEC linked with base station(BS).This work designs a task computation offloading model in heterogeneous vehicular network;this model takes into multiple stochastic tasks,the variety of wireless channels and bandwidth account.To obtain the tradeoff between the cost of energy consumption and the cost of data transmission delay and avoid curse of dimensionality caused by the complexity of the great action space,we propose an adaptive computation offloading method based on DRL that can address the continuous action space.The proposed algorithm adds the Ornstein-Uhlenbeck(OU)noise vector to the action space with different factors for each action to validate the exploration.Multiple transmission equipment can execute local processing and computation offloading to MEC.Nevertheless,ACORL considers the variety of wireless channels and available bandwidth between adjacent time slots.The numerical results illustrate that the proposed algorithm can effectively learn the optimal policy,which outperforms two baselines in the stochastic environment.Finally,the use of renewable energy harvesting capabilities in base stations or Internet of things(Io T)nodes may reduce energy consumption.As wireless channel conditions vary with time and the arrival rates of renewable energy,computing tasks are stochastic,and data offloading and renewable energy aware for Io T devices under a dynamic and unknown environment are major challenges.This work designs a data offloading and renewable energy aware model considering an MEC server performing multiple stochastic computing tasks and involving time-varied wireless channels.To optimize data transmission delay,energy consumption,and bandwidth allocation jointly,and to avoid the curse of dimensionality caused by the complexity of the state space,we propose a joint optimization method for data offloading,renewable energy aware,and bandwidth allocation for Io T devices based on DRL,which can handle the continuous action space.The proposed algorithm can minimize the long-term system cost(including data buffer delay cost,energy consumption cost,and bandwidth cost)and obtain an efficient solution by adaptively learning from the dynamic Io T environment.The numerical results verify the effectiveness and superiority of proposed algorithm compared to other baselines.