关键词： mobile edge computing   age of information   constrained Markov decision process   structural analysis   Q-learning  

摘要： Fresh status updates are vital to the efficient operation of network monitoring and real-time control applications. In this paper, we consider a mobile edge computing(MEC)-assisted status update system, where smart devices extract valuable status updates from sensed data to achieve timely awareness of the surroundings by exploiting computational resources at the device and edge server. To quantify the freshness of status updates obtained by executing computation tasks, we employ the concept of age of information(Ao I) to characterize the timeliness of status updates. To cope with the limited energy at devices, we investigate a joint task generation and computation offloading scheme under a given energy budget for minimizing the age of obtained status updates. The age minimization problem is modeled as a constrained Markov decision process(CMDP). To obtain the optimal policy, we derive the structural properties of the optimal deterministic policy and propose a lightweight structure-based status update algorithm in the case of known channel statistics. Moreover, we consider a more realistic scenario without prior knowledge of channel statistics, and propose a Q-learning-based status update algorithm to make online decisions. Simulation results show that the performance of our proposed algorithms is competitive when compared with existing schemes.

关键词： Optimization   Streaming media   Transcoding   Servers   Artificial neural networks   Edge computing   Resource management   Mobile edge computing   small cell networks   computation offloading   resource allocation   in-network caching  

摘要： In this correspondence, we jointly optimize the caching decision and radio and computation resource allocation for video service in mobile edge computing (MEC) enabled network to maximize the system revenue. To tolerate the uncertainty in the network traffic and avoid "hard constraint" based on constant content request rate, we employ robust optimization to obtain optimal caching decisions, and then accordingly allocate the radio and computation resources for video transcoding. Simulation results demonstrate the effectiveness of the proposed scheme.

关键词： Task analysis   Wireless communication   Computational modeling   Servers   Resource management   Edge computing   Optimization   Mobile edge computing   optimization algorithm   deep reinforcement learning   resource allocation  

摘要： In this paper, we consider a mobile-edge computing (MEC) system, where an access point (AP) assists a mobile device (MD) to execute an application consisting of multiple tasks following a general task call graph. The objective is to jointly determine the offloading decision of each task and the resource allocation (e.g., CPU computing power) under time-varying wireless fading channels and stochastic edge computing capability, so that the energy-time cost (ETC) of the MD is minimized. Solving the problem is particularly hard due to the combinatorial offloading decisions and the strong coupling among task executions under the general dependency model. Conventional numerical optimization methods are inefficient to solve such a problem, especially when the problem size is large. To address the issue, we propose a deep reinforcement learning (DRL) framework based on the actor-critic learning structure. In particular, the actor network utilizes a DNN to learn the optimal mapping from the input states (i.e., wireless channel gains and edge CPU frequency) to the binary offloading decision of each task. Meanwhile, by analyzing the structure of the optimal solution, we derive a low-complexity algorithm for the critic network to quickly evaluate the ETC performance of the offloading decisions output by the actor network. With the low-complexity critic network, we can quickly select the best offloading action and subsequently store the state-action pair in an experience replay memory as the training dataset to continuously improve the action generation DNN. To further reduce the complexity, we show that the optimal offloading decision exhibits an one-climb structure, which can be utilized to significantly reduce the search space of action generation. Numerical results show that for various types of task graphs, the proposed algorithm achieves up to 99.1% of the optimal performance while significantly reducing the computational complexity compared to the existing optimization me

关键词： Rendering (computer graphics)   Quality of experience   Servers   Wireless networks   Mobile handsets   Games   Wireless virtual reality   mobile edge computing   future wireless networks   offloading   and caching  

摘要： Wireless virtual reality (VR) is predicted to become a killer application in 5G and beyond, which provides an immersive experience and revolutionizes the way people communicate. It is well-known that rendering is the key performance bottleneck in wireless VR systems, especially for VR games. However, real-time rendering and data correlation are ignored by most researchers. In this paper, we propose an adaptive VR framework that enables high-quality wireless VR in future mmWave-enabled wireless networks with mobile edge computing (MEC), where real-time VR rendering tasks can be offloaded to MEC servers adaptively and the caching capability of MEC servers enables further performance improvement. First, we formulate the addressed problem to maximize the quality of experience (QoE) of the users, where association, adaptive offloading mode selection, and caching policy are jointly optimized. Considering the high complexity of the addressed problem, we then propose a distributed learning approach consisting of an offline training phase and an online running phase, which maintains scalability and adaptation capability. The offline phase is based on deep reinforcement learning (DRL) while the latter utilizes game theory. At last, simulation results show the superiority of the proposed algorithm over the other baseline algorithms in terms of QoE utility values, latency, and convergence time.

关键词： Accidents   Computational modeling   Task analysis   Delays   Data integrity   Data models   Autonomous vehicles   Autonomous driving   mobile edge computing   deep learning   computation offloading  

摘要： This letter proposes an edge learning-based offloading framework for autonomous driving, where the deep learning tasks can be offloaded to the edge server to improve the inference accuracy while meeting the latency constraint. Since the delay and the inference accuracy are incurred by wireless communications and computing, an optimization problem is formulated to maximize the inference accuracy subject to the offloading probability, the pre-braking probability, and data quality. Simulations demonstrate the superiority of the proposed offloading framework.

关键词： Task analysis   Energy consumption   Servers   Network security   Physical layer   Resource management   Delays   Computation offloading   energy consumption   mobile edge computing   physical layer security   resource allocation  

摘要： Mobile edge computing (MEC) has become a promising technology for real-time communications. Mobile devices can reduce the energy consumption and prolong the lifetime significantly via offloading the computing tasks to the MEC server. Moreover, physical layer security techniques can ensure the secure transmission of the offloading data. This paper investigates a MEC system that consists of an access point, multiple mobile devices and a malicious eavesdropper. The tasks allocation, local central processor's frequency, offloading power, and offloading timeslots are optimized jointly to minimize the total energy consumption of the system. A difference of convex algorithm based scheme is proposed to solve the joint optimization problem. Moreover, a Karush Kuhn Tucker conditions based algorithm is also proposed to reduce the computational complexity. Numerical results show that the proposed algorithms are very effective. Moreover, the power consumption for secure offloading decreases with the increase of the distance between the mobile devices and the eavesdropper.

关键词： Mobile edge computing   DASH   Layered encoding video   Delay analysis   Caching optimization  

摘要： Recently, the demand for video applications has become more and more pressing. To save backhaul bandwidth and reduce the network delay, Mobile Edge Computing (MEC) has arisen as a promising technique in future mobile networks. It can provide the storage, computing and networking resources in close-proximity to the mobile users through deploying multiple MEC servers at mobile edge. However, how to effectively utilize the limited storage capacity of these MEC servers is still the major challenge, especially when Dynamic Adaptive Streaming over HTTP (DASH) is used for video transmission. Recent studies have shown that layered encoding is very suitable for streaming service using DASH standard. Therefore, in our paper, combining with the feature of DASH-based scalable video transmission, we first design a novel multi-server cooperative scheme with software-defined networking (SDN) architecture for future MEC-assisted heterogeneous networks, to effectively manage and schedule the available resources, thus guaranteeing the requested video delivered to the clients with higher quality and lower delay. And then, based on the above scheme, the end-end delay is carefully analyzed for each real video streaming. With a full consideration of the averag end-to-end delay of video streaming and the provided video quality of MEC servers, we finally propose a Dynamic Programming based Adaptive Caching Algorithm (DPACA) to obtain an optimized caching strategy under the constraint of cache storage capacity at each MEC sever. Simulation results show that, our algorithm can improve the cache hit ratio while significantly reducing the end-to-end delay. Moreover, our algorithm has higher flexibility and adaptability.

关键词： radio access networks   cloud computing   cellular radio   mobile computing   resource allocation   virtual machines   convergence   optimisation   joint optimisation problem   separate resource allocation   communication resources   virtual machine selection algorithm   multihop cloud   mobile edge computing   mobile devices   base station   wireless resources   high transmission delays   multihop path   edge computing resources   cloud radio access network   wireless self-organising network   task delay   elite selection   improved krill herd algorithm  

摘要： For the problem that mobile devices that are far away from the base station obtain limited wireless resources, which causes high transmission delays, the study proposes a multi-hop path to assist users who are far away from the base station to use edge computing resources. And it combines cloud radio access network and wireless self-organising network. The modelling aims to minimise the total energy consumption under the constraints of task delay considering the mobility of mobile devices. Meanwhile, it selects the joint optimisation problem of the path for data transmission and the virtual machine for calculation. This study also introduces the krill herd algorithm and analyses its advantages and disadvantages. The author enhances the global search ability of the algorithm by defining perturbation factors in the random diffusion behaviour and introduces a new strategy of elite selection and retention into the iterative process to improve the convergence accuracy. Finally, the improved krill herd algorithm is used to solve the joint optimisation problem and a better allocation result than the separate resource allocation of calculation and communication is obtained. The experiment proves that the selection algorithm combining virtual machine and routing proposed in this study can achieve the expected results.

关键词： Reinforcement learning   Actor-critic algorithm   Eligibility traces   Mobile edge computing   Resource allocation  

摘要： Reinforcement learning (RL) as an effective tool has attracted great attention in wireless communication field nowadays. In this paper, we investigate the offloading decision and resource allocation problem in mobile edge computing (MEC) systems based on RL methods. Different from existing literature, our research focuses on improving mobile operators' revenue by maximizing the amount of the offloaded tasks while decreasing the energy expenditure and time-delays. Considering the dynamic characteristics of wireless environment, the above problem is modeled as a Markov decision process (MDP). Since the action space of the MDP is multidimensional continuous variables mixed with discrete variables, traditional RL algorithms are powerless. Therefore, an actor-critic (AC) with eligibility traces algorithm is proposed to resolve the problem. The actor part introduces the parameterized normal distribution to generate the probabilities of continuous stochastic actions, and the critic part employs a linear approximator to estimate the value of states, based on which the actor part updates policy parameters in the direction of performance improvement. Furthermore, an advantage function is designed to reduce the variance of the learning process. Simulation results indicate that the proposed algorithm can find the best strategy to maximize the amount of the tasks executed by the MEC server while decreasing the energy consumption and time-delays.

关键词： Servers   Task analysis   Mobile handsets   Delays   Computational modeling   Processor scheduling   Energy consumption   Conflict graphs   genetic algorithms   mobile edge computing   parallel offloading   sequential offloading  

摘要： In this paper, we consider parallel and sequential task offloading to multiple mobile edge computing servers. The task consists of a set of inter-dependent sub-tasks, which are scheduled to servers to minimize both offloading latency and failure probability. Two algorithms are proposed to solve the scheduling problem, which are based on genetic algorithm and conflict graph models, respectively. Simulation results show that these algorithms provide performance close to the optimal solution, which is obtained through exhaustive search. Furthermore, although parallel offloading uses orthogonal channels, results demonstrate that the sequential offloading yields a reduced offloading failure probability when compared to the parallel offloading. On the other hand, parallel offloading provides less latency. However, as the dependency among sub-tasks increases, the latency gap between parallel and sequential schemes decreases.