关键词： 5G   communication overhead   DDPG   mobile edge computing   reinforcement learning   resource allocation  

摘要： Distributed base station deployment, limited server resources and dynamically changing end users in mobile edge networks make the design of computing offloading schemes extremely challenging. Considering the advantages of deep reinforcement learning (DRL) in dealing with dynamic complex problems, this paper designs an optimal computing offloading and resource allocation strategy. Firstly, the authors consider a multi-user mobile edge network scenario consisting of Macro-cell Base Station (MBS), Small-cell Base Station (SBS) and multiple terminal devices, the communication overhead and calculation overhead generated are formulated and described in detail. Besides, combined with the deterministic delay of tasks, the optimization objective of this paper is clarified to comprehensively consider system energy consumption. Then, a learning algorithm based on Deep Deterministic Policy Gradient (DDPG) is proposed to minimize system energy consumption. Finally, simulation experiments show that the authors' proposed DDPG algorithm can effectively optimize the target value, and the total system energy consumption is only 15.6 J, which is better than other compared algorithms. It is also proved that the proposed algorithm has excellent communication resource allocation ability.

关键词： mobile edge computing   unmanned aerial vehicle   task migration   deep reinforcement learning   federated learning  

摘要： At present, with the intelligence that has been achieved in computer and communication technologies, vehicles can provide many convenient functions to users. However, it is difficult for a vehicle to deal with computationally intensive and latency-sensitive tasks occurring in the vehicle environment by itself. To this end, mobile edge computing (MEC) services have emerged. However, MEC servers (MECSs), which are fixed on the ground, cannot flexibly respond to temporal dynamics where tasks are temporarily increasing, such as commuting time. Therefore, research has examined the provision of edge services using additional unmanned aerial vehicles (UAV) with mobility. Since these UAVs have limited energy and computing power, it is more important to optimize energy efficiency through load balancing than it is for ground MEC servers (MECSs). Moreover, if only certain servers run out of energy, the service coverage of a MEC server (MECS) may be limited. Therefore, all UAV MEC servers (UAV MECSs) need to use energy evenly. Further, in a high-mobility vehicle environment, it is necessary to have effective task migration because the UAV MECS that provides services to the vehicle changes rapidly. Therefore, in this paper, a federated deep Q-network (DQN)-based task migration strategy that considers the load deviation and energy deviation among UAV MECSs is proposed. DQN is used to create a local model for migration optimization for each of the UAV MECSs, and federated learning creates a more effective global model based on the fact that it has common spatial features between adjacent regions. To evaluate the performance of the proposed strategy, the performance is analyzed in terms of delay constraint satisfaction, load deviation, and energy deviation.

关键词： Mobile edge computing   Multi-user service migration   Deep reinforcement learning   Partially observable Markov decision processes  

摘要： With the continuous development of mobile edge computing, people are more willing to offload tasks to edge servers that are closer to users than cloud services for a better user experience. Due to problems such as user mobility and limited coverage of edge servers, how to ensure service quality and prevent service interruption is a problem we have to consider. This article explores the problem of service migration to decide when, where, and how to migrate an ongoing service from the current edge server to the target edge server. At the same time, we will consider the premise that edge servers can only obtain partial user information, and model the service migration problem as a partially observable Markov decision process. To minimize user delay and system energy consumption, we propose a service migration decision algorithm based on Deep recurrent Q-learning (DRQNSM). Multiple experiments show that our algorithm has a better performance than some classical reinforcement learning algorithms.

关键词： mobile edge computing   task offloading   security   unmanned aerial vehicle   compression  

摘要： Nowadays, Unmanned Aerial Vehicle (UAV) devices and their services and applications are gaining popularity and attracting considerable attention in different fields of our daily life. Nevertheless, most of these applications and services require more powerful computational resources and energy, and their limited battery capacity and processing power make it difficult to run them on a single device. Edge-Cloud Computing (ECC) is emerging as a new paradigm to cope with the challenges of these applications, which moves computing resources to the edge of the network and remote cloud, thereby alleviating the overhead through task offloading. Even though ECC offers substantial benefits for these devices, the limited bandwidth condition in the case of simultaneous offloading via the same channel with increasing data transmission of these applications has not been adequately addressed. Moreover, protecting the data through transmission remains a significant concern that still needs to be addressed. Therefore, in this paper, to bypass the limited bandwidth and address the potential security threats challenge, a new compression, security, and energy-aware task offloading framework is proposed for the ECC system environment. Specifically, we first introduce an efficient layer of compression to smartly reduce the transmission data over the channel. In addition, to address the security issue, a new layer of security based on an Advanced Encryption Standard (AES) cryptographic technique is presented to protect offloaded and sensitive data from different vulnerabilities. Subsequently, task offloading, data compression, and security are jointly formulated as a mixed integer problem whose objective is to reduce the overall energy of the system under latency constraints. Finally, simulation results reveal that our model is scalable and can cause a significant reduction in energy consumption (i.e., 19%, 18%, 21%, 14.5%, 13.1% and 12%) with respect to other benchmarks (i.e., local, edg

关键词： Software-defined networking   mobile edge computing   5G   approximation algorithm  

摘要： The software-defined network (SDN) enabled mobile edge network greatly facilitates network resource management and promotes many emerging applications. However, user mobility may cause the SDN controller to set flow rules frequently, introduce additional flow setup latency, cause delay jitter, and undermine latency-sensitive services. Proactive flow setup is an effective way to eliminate flow setup latency, but existing work fails to maximize the flow setup hit ratio, a metric for evaluating the quality of proactive flow setup decisions, which is critical for latency-sensitive services. In this paper, we study how to proactively set flow rules to maximize the flow setup hit ratio under limited available network resources to eliminate the flow setup latency as much as possible. Then, we formalize the proactive flow setup problem as two integer linear programming problems under two typical routing strategies, default routing and dynamic routing. Both problems are proved to be NP-hard. To tackle these two problems, we propose a linear programmingbased polynomial-time approximation algorithm for the default routing case and a greedy-based heuristic algorithm for the dynamic routing case. Extensive trace-driven experimental and simulation results verify that our algorithms can improve the flow setup hit ratio by up to 30.99% compared to existing solutions.

关键词： Computation offloading   Downlink and uplink decoupling   Heterogeneous networks   Mobile edge computing   Resource allocation  

摘要： Mobile edge computing (MEC) is a key player in low latency 5G networks with the task to resolve the conflict between computationally-intensive mobile applications and resource-limited mobile devices (MDs). As such, there has been intense interest in this topic, especially in multi-user single-server and homogeneous multi-server scenarios. However, the research in the heterogeneous multi-server scenario is limited, where the servers are located at small base-stations (SBSs), macro base-stations (MBSs), or the cloud with different computing and communication capabilities. On the other hand, computational-tasks offloading is limited by the type of MD-BS association with almost all previous works focusing on offloading the MD's computational tasks to the MEC servers/cloudlets at its serving BS. However, in multi-BS association, or downlink/uplink decoupled (DUDe) scenarios, an MD can be served by multiple BSs and hence has multiple offloading choices. Motivated by this, we proposed a joint BS association and subchannel allocation algorithm based on a student-project allocation (SPA) matching approach to minimize the network sum-latency, which breaks the constraint that one MD must connect to the same BS in the UL and DL, and jointly consider the communication and computational disparity of SBS and MBS cloudlets in heterogeneous MEC networks. Moreover, an optimal power allocation scheme is proposed to optimize the system performance subject to the predefined quality of service constraints. Our results show that the proposed scheme is superior to benchmark techniques in enabling the effective use of computational and communication resources in heterogeneous MEC networks.

关键词： Industrial Internet of Things   mobile edge computing   wireless power transfer   heterogeneous computing tasks   resource allocation  

摘要： Integrating wireless power transfer with mobile edge computing (MEC) has become a powerful solution for increasingly complicated and dynamic industrial Internet of Things (IIOT) systems. However, the traditional approaches overlooked the heterogeneity of the tasks and the dynamic arrival of energy in wirelessly powered MEC-enabled IIOT systems. In this article, we formulate the problem of maximizing the product of the computing rate and the task execution success rate for heterogeneous tasks. To manage energy harvesting adaptively and select appropriate computing modes, we devise an online resource allocation and computation offloading approach based on deep reinforcement learning. We decompose this approach into two stages: an offloading decision stage and a stopping decision stage. The purpose of the offloading decision stage is to select the computing mode and dynamically allocate the computation round length for each task after learning from the channel state information and the task experience. This stage allows the system to support heterogeneous computing tasks. Subsequently, in the second stage, we adaptively adjust the number of fading slots devoted to energy harvesting in each round in accordance with the status of each fading slot. Simulation results show that our proposed algorithm can better allocate resources for heterogeneous tasks and reduce the ratio of failed tasks and energy consumption when compared with several existing algorithms.

关键词： Mobile edge computing   intelligent reflecting surface   wireless power transfer   energy beamforming   multiple-user detection  

摘要： This paper proposes a multiple intelligent reflecting surfaces (IRSs) assisted wireless-powered mobile edge computing (MEC) system, where the IRSs are deployed to assist both the downlink wireless power transfer (WPT) from the multi-antenna hybrid access point (HAP) to the wireless devices (WDs) and the uplink computation offloading from the WDs to the MEC server. To further improve the system performance, the energy beamforming and multiple-user detection (MUD) technologies are exploited. We consider both partial and binary offloading schemes and formulate the sum computation rate (SCR) maximization problems for them, respectively. To tackle the non-convexity of each problem, we propose an efficient alternating optimization (AO) method. Specifically, the Lagrange duality method is used to optimize the energy beamforming vector and the MUD matrix at the HAP, and the CPU frequencies and transmit power of the WDs. Then, we optimize the discrete phase shifts via the successive convex appropriation (SCA) method, the rank-one equivalents, and rounding method. Finally, the optimal time scheduling can be obtained via the one-dimensional search method. In addition, we propose a greedy algorithm with low complexity to optimize the computing modes for the binary offloading scheme. Numerical results show that our proposed schemes outperform the benchmarks.

关键词： Roads   Servers   Task analysis   Computer architecture   Heuristic algorithms   Optimization   Image edge detection   Convex optimization   deep reinforcement learning   latency optimization   mobile edge computing   proximity detection  

摘要： In road networks, mobile users (including vehicles and pedestrians) need to know the proximity relationship with other users in real time, referred to as the problem of proximity detection which is very significant for autonomous driving. However, due to limited computing and storage resources of mobile users and real-time changes of road network status, it becomes a difficult task to calculate and update the proximity relationship between users in real time. Therefore, in this paper, we first propose a computation offloading scheme and a dynamic road network state update model for proximity detection in dynamic road networks based on Mobile Edge Computing (MEC), and formulate the latency optimization problem for proximity detection in the dynamic road network as a nonlinear optimization problem. Then we use the Sequential Least Squares Programming (SLSQP) algorithm to solve the latency optimization problem. In addition, to reduce the running time, we also use the deep reinforcement learning approach, i.e., the Deep Deterministic Policy Gradient (DDPG) algorithm, to address the latency optimization problem. Simulation results show that, compared with the SLSQP algorithm, the DDPG algorithm can effectively and efficiently reduce the computational time of the optimal latency each time by continuously adjusting the task allocation strategy, and the computational time of the DDPG algorithm is two orders of magnitude lower than the SLSQP algorithm.

关键词： Task analysis   Servers   Relay networks (telecommunication)   Resource management   Computational modeling   Delays   Costs   Caching   mobile edge computing   stable matching   task offloading   ultra-dense relay network  

摘要： In recent years, ultra-dense relay network (UDRN) has become a hot research topic. Mobile edge computing (MEC) can improve the performance of intelligent interactive applications in terms of time delay and energy consumption. In traditional MEC, MEC server in the base station (BS) is responsible for all the computing in the network. However, the MEC deployed in the BS may lead to huge computation burden on the BS and heavy traffic load through the network. Deploying MEC server in the relays of the network, which helps computing and caching for the users, will lower the burden of the BS, reduce the demand for traffic transmission through the network and improve the quality of service for users. This paper studies the problem of task caching, offloading and stable matching (SM) in UDRNs with MEC deployed in the relays, which is formulated as a multi-variable integer programming problem. To enable task caching and offloading, we propose an improved branch-and-bound algorithm and a low-complexity suboptimal algorithm to obtain the optimal and suboptimal schemes, respectively. Then, according to the principle of minimizing individual costs, the mutual preference lists between the users and relays are established. Based on the mutual preference lists, we propose an improved SM algorithm to obtain the user-relay SM results. Simulation results show that the proposed algorithms can bring considerable performance gains compared with conventional algorithms.