The proliferation of smart vehicular terminals (VTs) and their resource hungry applications imposes serious challenges to the processing capabilities of VTs and the delivery of vehicular services. Mobile Edge Computing (MEC) offers a promising paradigm to solve this problem by offloading VT applications to proximal MEC servers, while TV white space (TVWS) bands can be used to supplement the bandwidth for computation offloading. In this paper, we consider a cognitive vehicular network (CVN) that uses the TVWS band, and formulate a dual-side optimization problem, to minimize the cost of VTs and that of the MEC server at the same time. Specifically, the dual-side cost minimization is achieved by jointly optimizing the offloading decision and local CPU frequency on the VT side, and the radio resource allocation and server provisioning on the server side, while guaranteeing network stability. Based on Lyapunov optimization, we design an algorithm called DDORV to tackle the joint optimization problem, where only current system states, such as channel states and traffic arrivals, are needed. The closed-form solution to the VT-side problem is obtained easily by derivation and comparing two values. For MEC server side optimization, we first obtain server provisioning independently, and then devise a continuous relaxation and Lagrangian dual decomposition based iterative algorithm for joint radio resource and power allocation. Simulation results demonstrate that DDORV converges fast, can balance the cost-delay tradeoff flexibly, and can obtain more performance gains in cost reduction and as compared with existing schemes.

Immersive virtual reality (VR) video is becoming increasingly popular owing to its enhanced immersive experience. To enjoy ultra-high resolution immersive VR video with wireless user equipments such as head-mounted displays (HMDs), ultra-low-latency viewport rendering and data transmission are the core prerequisites, which could not be achieved without a huge bandwidth and superior processing capabilities. Besides, potentially very high energy consumption at the HMD may impede the rapid development of wireless panoramic VR video. Multi-access edge computing (MEC) has emerged as a promising technology to reduce both the task processing latency and the energy consumption for HMD, while bandwidth-rich THz communication is expected to enable ultra-high-speed wireless data transmission. In this paper, we propose to minimize the long-term energy consumption of a THz wireless access based MEC system for high quality immersive VR video services support by jointly optimizing the viewport rendering offloading and downlink transmit power control. Considering the time-varying nature of wireless channel conditions, we propose a deep reinforcement learning based approach to learn the optimal viewport rendering offloading and transmit power control policies and an asynchronous advantage actor-critic (A3C) based joint optomization algorithm is proposed. Simulation results demonstrate that the proposed algorithm converges fast under different learning rates, and outperforms existing algorithms in terms of minimized energy consumption and maximized reward.

When using blockchain in mobile systems, computation intensive mining tasks pose great challenges to the processing capabilities of mobile miner equipment. Mobile edge computing (MEC) is an effective solution to alleviating the problem via task offloading. In the mining process, miners compete for rewards through puzzle solving, where only the miner that first completes the process will be rewarded. Thus, miners may wish to pay higher price and use more communication resources in task offloading and more computation resources in task processing for latency reduction. However, there are risks for the miners not profiting from consuming more resources or paying a higher price, so miners are rational in blockchain systems. In order to maximize the rational total profit of all miners, we use an asynchronous advantage actor-critic (A3C) deep reinforcement learning algorithm to obtain the resource pricing and allocation, considering the stochastic properties of wireless channels, and the prospect theory is employed to strike a good balance between risks and rewards. Numerical results show that our proposed A3C based joint optimization algorithm converges fast and outperforms the baseline algorithms in terms of the total reward.

Ambient backscatter communications (AmBackComs) have been recognized as a spectrum-and energy-efficient technology for Internet of Things, as it allows passive backscatter devices (BDs) to modulate their information into the legacy signals, e.g., cellular signals, and reflect them to their associated receivers while harvesting energy from the legacy signals to power their circuit operation. However, the co-channel interference between the backscatter link and the legacy link and the non-linear behavior of energy harvesters at the BDs have largely been ignored in the performance analysis of AmBackComs. Taking these two aspects, this paper provides a comprehensive outage performance analysis for an AmBackCom system with multiple backscatter links, where one of the backscatter links is opportunistically selected to leverage the legacy signals transmitted in a given resource block. For any selected backscatter link, we propose an adaptive reflection coefficient (RC), which is adapted to the non-linear energy harvesting (EH) model and the location of the selected backscatter link, to minimize the outage probability of the backscatter link. In order to study the impact of co-channel interference on both backscatter and legacy links, for a selected backscatter link, we derive the outage probabilities for the legacy link and the backscatter link. Furthermore, we study the best and worst outage performances for the backscatter system where the selected backscatter link maximizes or minimizes the signal-to-interference-plus noise ratio (SINR) at the backscatter receiver. We also study the best and worst outage performances for the legacy link where the selected backscatter link results in the lowest and highest co-channel interference to the legacy receiver, respectively. Computer simulations validate our analytical results, and reveal the impacts of the co-channel interference and the EH model on the AmBackCom performance. In particular, the co-channel interference leads to the outage saturation phenomenon in AmBackComs, and the conventional linear EH model results in an over-estimated outage performance for the backscatter link.

Emerging smart Internet-of-Things (IoT) applications are increasingly relying on mobile-edge computing (MEC) networks, where the energy efficiency (EE) of computation is one of the most pertaining issues. In this article, considering the limited computation capacity at the MEC server and a practical nonlinear energy harvesting (EH) model for IoT devices, we propose a scheme to maximize the system computation EE (CEE) of a wireless power transfer (WPT) enabled nonorthogonal multiple access (NOMA)-based MEC network by jointly optimizing the computing frequencies and execution time of the MEC server and the IoT devices, the offloading time, the EH time and the transmit power of each IoT device, as well as the transmit power of the power beacon (PB). We formulate the joint optimization into a nonlinear fractional programming problem and devise a Dinkelbach-based iterative algorithm to solve it. By means of convex theory, we derive closed-form expressions for parts of the optimal solutions, which reveal several instrumental insights into the maximization of the system CEE. In particular, the system CEE increases as the optimal computing frequencies of both the IoT devices and the MEC server decrease, and the system CEE is maximized when the MEC server and the IoT devices use the maximum allowed time to complete their computing tasks. Simulation results demonstrate the superiority of the proposed scheme over benchmark schemes in terms of system CEE.

In this paper, we consider a backscatter-assisted wireless powered mobile edge computing (MEC) network, where multiple Internet-of-Things (IoT) nodes harvest energy from the energy signals transmitted by a power beacon (PB) and utilize the harvested energy for local computing and task offloading via hybrid backscatter communication (BackCom) and active transmission (AT). Considering the limited computation capacity of the MEC server and the quality-of-service (QoS) and energy-causality constraints per IoT node, we propose two resource allocation schemes to maximize the total computation bits of all the IoT nodes and the system computation energy efficiency (EE), respectively, by jointly optimizing the computation frequency and time of the MEC server and each IoT node, the transmit power of the PB and each IoT node, and the BackCom power reflection coefficient and the time for energy harvesting (EH), BackCom, and AT of each IoT node. The non-convex computation bits maximization problem is transformed to a convex one by introducing a series of auxiliary variables and proof by contradiction, and then solved by the existing convex tools. The system computation EE maximization is a non-convex nonlinear programming problem. We propose a two-layer iterative algorithm to solve it optimally and devise a reduced-complexity iterative algorithm to solve it sub-optimally by leveraging the block coordinate decent technique. Computer simulations validate the convergence of the proposed iterative algorithms and their superior performance over the benchmark schemes in terms of the computation bits or EE.