The coronavirus disease 2019 (COVID-19) has recently attracted extensive attention due to its serious impact on public health worldwide. In this paper, we study and verify that the popularity of virus-related content has a negative correlation with the epidemic spread by means of statistical analysis. Inspired by this result, a practical solution of recommender system is proposed for pushing virus-related content, aiming to gain insight about the newly discovered virus for people and thus reduce the epidemic spread to the utmost extent. First, we formulate the optimization of recommendation policy subject to quality of experience (QoE) loss constraints as a finite-horizon Constrained Markov Decision Problem (CMDP). To solve this problem, then, we present both enumeration and heuristic methods, from perspectives of achieving optimal recommendation policy and reducing computational complexity, respectively. Finally, our simulations validate the benefit of our solution by showing that to recommend virus-related content following our strategy does help slow down the spread of the epidemic.

In a vehicular network environment, vehicles can download data through opportunistically-encountered Roadside Units (RSUs) with a lower cost, compared to that from Base Stations (BSs). However, the delay experienced by the vehicles might be undesirably prolonged if they only download data through RSUs. In this paper, we aim to minimize the average delay under the constraint of average cost by scheduling the download rates from RSUs and BSs. One challenge lying in the design of the downloading policy is the uncertainty of the download condition, i.e., whether at least an RSU is available or BS only, in the future slots. To overcome this challenge, we notice that vehicles from opposite directions can share their known download conditions to reduce this uncertainty. To make the most of this information, a Markov decision process (MDP) is used to model the system operations, based on which, average delay and cost can be analyzed to formulate the optimization problem. By solving this problem, the delay-minimal downloading policy can be obtained to achieve the optimal delay-cost tradeoff in the considered vehicular network. Finally, performance improvement with the help of information sharing among vehicles is validated by extensive simulations.

Wireless power transfer (WPT) is an effective approach to enhance sustainability of mobile devices. In this context, we consider a system composed of a hybrid access point (HAP) that powers a wireless device (WD), provisioned with a finite energy storage and data butter. The WD uses the harvested energy to perform data transmission. We assume that both HAP and WD can perform power adaptation to save energy. Unlike most existing works, we focus on delay minimization in transmitting an arbitrary data arrival packets over wireless fading channels, which is extremely important for time-sensitive applications. Previous works mainly focused on slot-oriented optimization, where the harvested energy is consumed in the same slot, without considering the possibility of storing energy for future use. In contrast, we consider a long-term average delay minimization under average power consumption constraint at the HAP. This requires to consider the data arrival process, data and energy queue evolution, and the channel state statistics, thus, greatly increasing the optimization complexity. Our goal is to determine the optimal delay-power tradeoff and its corresponding scheduling strategy to decide the operations, e.g., data transmission by the WD and power transfer by the HAP, for arbitrary i.i.d arrival process and adaptive transmissions. Through two-dimensional Markov chain modeling and linear programming, we provide the optimal scheduling strategy. We corroborate the theoretical analysis by simulation results.

Mobile edge computing (MEC) is a promising technique to improve the quality of computation experience for mobile devices by providing computation resources in their close proximity. However, the design of scheduling policies for MEC systems inevitably encounters a challenging optimization problem that should take both transmissions and computations into consideration. In particular, how to jointly schedule transmissions and computations should adapt to the cross-layer system dynamics, i.e., random task arrivals and channel state variations. We formulate this scheduling problem as a joint optimization problem for both transmissions and computations in order to minimize the power consumption of mobile devices, while meeting the latency requirement. With given distributions of the system dynamics, Markov decision process (MDP) is used to model the system operations. Based on this model, the power-optimal scheduling policy can be obtained by converting the joint optimization problem to linear programming (LP) by using variable substitutions and thus the optimal power-latency tradeoff can be achieved. When the distribution information of the system dynamics is unknown, we exploit the Lyapunov optimization to present a low complexity scheduling policy. Our theoretical analysis and extensive simulation studies show that our approach can offer a good tradeoff between power consumption and latency.

Mobile cloud computing systems, or simply mobile clouds, have attracted tremendous attention because they allow mobile devices with limited computational resources to offload complex computations. However, due to the channel uncertainty and the complexity of a computation task, mobile computation offloading may suffer from poor outage performance that the offloaded task cannot be completed within the desired delay constraint. Thus, how to efficiently identify and overcome the outage bottleneck, which could be used to optimize resource allocation schemes and improve the system performance effectively is an open problem. In this paper, we shall develop a unified framework that minimizes the overall outage probability in various mobile computation offloading scenarios. More specifically, the outage bottleneck is defined and identified by adopting asymptotic analysis, without any need of the accurate outage probabilities in both transmissions and computations. To overcome the outage bottleneck, resource pairing, matching, and allocation policies are investigated. Both theoretical analysis and numerical results show that the outage bottleneck relies on not only the availability of spectrum and computation resources but also the probability distributions of computation complexities of the computation tasks.

A typical application of unmanned aerial vehicles (UAVs) is surveillance of distant targets, where data collected by its sensors need to be transmitted back to a ground terminal (GT) for further processing in a timely manner. Due to the limited battery capability of the UAV, the sensed data could be preprocessed in a UAV to reduce the amount of data transmitted, which could potentially reduce the average power consumption at the UAV, especially when the transmission link quality is poor. In this paper, a probabilistic approach is adopted to schedule the transmission and computing of the data tasks based on the UAV and GT's buffer states. The joint transmission and computing problem can be modeled as a four-dimensional Markov chain, based on which the average delay of each task and the average power consumption at the UAV can be obtained. Our design goal is to minimize the average power consumption under the delay constraints. To do that, a delay-constrained power minimization problem is solved by an proposed method to obtain the power-optimal joint transmission and computation scheduling (JTCS) policy efficiently. Finally, the optimization results are validated with extensive simulations.

In this paper, we aim to obtain the optimal tradeoff among average delay, and average transmission and computation power consumptions in a mobile computation offloading system. A probabilistic approach is developed to jointly determine the transmission and computation rate in each time-slot. We model the queue lengths in the mobile device and computation resource with a two- dimensional Markov chain. Based on this model, we obtain the average delay and power consumption. Then, we formulate a joint queues aware optimization problem to minimize the average power consumption of the mobile device given constraints on average delay of tasks and average power consumption of the computation resource. By converting the problem into a linear programming, we obtain the optimal power-delay tradeoff and power-optimal Joint Transmission and Computing Scheduling (JTCS) strategy. Finally, the optimization results are validated by extensive simulations.

Wireless cloud computing system with computation offloading has attracted much attention due to the potential of alleviating the restrictions of limited resources in mobile devices. During the computation offloading, the steps of transmissions and computation need to be completed within the required time, otherwise the system will be in outage. Due to uncertainties of channel fading and computational complexities, such outage is the result of either transmission outage or computation outage. In this paper, a theoretical framework is proposed to analyze the outage of the wireless cloud computing system with multiple subcarriers and computation resources. Through theoretical analysis of the outage probability, the outage bottleneck can be identified. Numerical results have verified the theoretical analysis and concluded that the outage bottleneck depends not only on the distributions of complexities of tasks but also on the numbers of subcarriers and computation resources.

Cooperative communication is a promising technique, which can greatly improve the performance and extend the coverage of cellular networks. In this paper, the resource allocation problem is investigated for multi-carrier amplify-and-forward (AF) relay assisted multi-access networks with multiple users and subcarriers. The joint relay selection and subcarrier allocation problem is formulated as a combinatorial optimization problem so as to minimize the outage probability with fairness assurance. To solve this problem, we propose a constraint maximum H-matching approach, which is based on a correlated random bipartite graph (RBG) formulation. By analyzing the properties of H-matching method on correlated RBG, the tight approximation of outage probability will be obtained in closed-form formulas. By deriving the diversity-multiplexing tradeoff, it is shown that the proposed approach achieves the same frequency and cooperative diversity as if the network has only one user, while all of the users fairly share the multiplexing gain at the same time. The proposed H-matching method also enjoys a sublinear computation complexity for parallel implementations. Numerical results verified the theoretical derivations and optimality of the proposed approach.