This article proposes a reinforcement-learning (RL) approach for optimizing charging scheduling and pricing strategies that maximize the system objective of public electric vehicle (EV) station. The proposed algorithm is "online" in sense decisions made at each time depend only on observation past events, "model-free" does not rely any assumed stochastic models uncertain events. To cope with challenge arising from time-varying continuous state action spaces RL problem, we first show it...

State estimation is critical to the monitoring and control of smart grids. Recently, false data injection attack (FDIA) emerging as a severe threat state estimation. Conventional FDIA detection approaches are limited by their strong statistical knowledge assumptions, complexity, hardware cost. Moreover, most current focus on detecting presence FDIA, while important information exact locations not attainable. Inspired recent advances in deep learning, we propose deep-learning-based locational...

The rapid emergence of electric vehicles (EVs) demands an advanced infrastructure publicly accessible charging stations that provide efficient services. In this paper, we propose a new station operation mechanism, the Joint Admission and Pricing (JoAP), which jointly optimizes EV admission control, pricing, scheduling to maximize station's profit. More specifically, by introducing tandem queueing network model, analytically characterize average profit as function control pricing policies....

In this paper, we consider the profit-maximizing demand response of an energy load in real-time electricity market. a market, market clearing price is determined by random deviation actual power supply and from predicted values day-ahead An load, which requires total amount over certain period time, has flexibility shifting its usage therefore perfect position to exploit volatile through response. We show that strategy can be obtained solving finite-horizon Markov decision process (MDP)...

This paper proposes a new model of scenario-based security-constrained unit commitment (SCUC) with BESSs. By formulating such as mixed-integer programming (MIP) problem, we can obtain the optimal control strategy units and BESSs to reduce operating cost. To solve this MIP proposed model, propose learning-based approach tackle SCUC problem. The convolutional neural network (CNN)-based algorithm (CNN-SCUC) has two main stages. First, CNN-SCUC trains CNN solutions binary variables corresponding...

The emerging mobile edge computing (MEC) technology has been recently applied to improve the Quality of Experience (QoE) network services, such as live video streaming. In this paper, we study an energy-aware adaptive streaming scheme in wireless networks. particular, aim design a joint uplink transmission and transcoding algorithm maximizing followers' QoE, while minimizing energy consumption streamer. We formulate problem Markov decision process (MDP), propose deep reinforcement learning...

With the advancement of multimedia technology and wireless networks, there is a growing demand for high-quality video streaming. Delivering stable streaming in extremely dynamic nevertheless, still an open problem. Recent developments client computing mobile edge (MEC) technologies have both shown promise enhancing adaptive bitrate (ABR) services. In this paper, we consider system multi-tier enabled by joint edge-side transcoding client-side enhancement. By “enhancement,” mean that improves...

The emerging mobile edge computing (MEC) technology has been recently applied to improve adaptive bitrate (ABR) streaming service quality under time-varying wireless channels. In this paper, we consider a heterogeneous multi-user MEC-enabled video network with channels in sequential time frames. particular, aim design an online joint transcoding and transmission resource allocation algorithm maximize the ABR user's of experience (QoE) subject bandwidth CPU constraints. is "online" sense that...

Electric vehicles (EVs) are regarded as one of the most effective ways to reduce oil demands and gas emissions. A charging station with well-designed scheduling pricing strategy can benefit both EV users electricity distribution system. This paper proposes a profit-maximizing joint scheme for public station. We first show that is solution Markov decision process (MDP). To solve problem, we propose reinforcement learning (RL) approach does not require any non-causal information or...

The emerging multi-access edge computing (MEC) technology effectively enhances the wireless streaming performance of 360-degree videos. By connecting a user's head-mounted device (HMD) to smart MEC platform, server (ES) can efficiently perform adaptive tile-based video improve viewing experience. Under constrained channel capacity, ES predict field view (FoV) and transmit HMD high-resolution tiles only within predicted FoV. In practice, is challenged by random FoV prediction error fading...

The rapid development of multimedia and communication technology has resulted in an urgent need for high-quality video streaming. However, robust streaming under fluctuating network conditions heterogeneous client computing capabilities remains a challenge. In this paper, we consider enhancement-enabled time-varying wireless limited computation capacity. "Enhancement" means that the can improve quality downloaded segments via image processing modules. We aim to design joint bitrate...

With growing popularity of enormous public safety and transportation infrastructure cameras, there are increasing demands for automatic mobile video analytics. The emerging multi-access edge computing (MEC) technology has been recently applied to improve the accuracy-latency tradeoff In this paper, we study an MEC-enabled multi-device analytics system formulate problem as a Markov decision process (MDP) meet two practical challenges: i) absence ground truth in real-time ii) content-varying...

Beam alignment is essential for high-quality data transmission in millimeter wave (mmWave) communication systems. Recent studies have revealed that the beam method can train a small-scale probing codebook customized to specific sites, and leverage measurements determine optimal transmit beam. However, existing approaches still necessitate certain number of beams order achieve high-accuracy alignment. In this work, we propose multi-task learning-based method, leveraging channel reconstruction...

The Internet of Things (IoT) technology connects various aspects society and enhances human life. Electric vehicles (EVs) charging stations (CSs) are essential components the IoT system, offering business opportunities for companies like aggregator. aggregator can maximize profit by implementing CSs management strategies, such as pricing scheduling. However, managing presents challenges due to irrational behavior, particularly uncertain CS selections EV users. To address this issue, we...

With the fast development of multi-sensor technologies, image fusion has played an essential role in modern military and civilian. To better integrate thermal radiation information infrared images detailed appearance visible images, we investigate a novel norm formulation via joint contrast gradient preservation, for infrared-visible fusion. Specifically, employ structure tensor measurement to characterize similarity between fused terms information, details. Since natural gradients follow...

Under the outbreak of COVID-19 pandemic, respiration state monitoring plays an important role in assisting respiratory disease diagnosis and treatment. Thanks to nonintrusive nature low deployment cost, Wi-Fi-based wireless methods have gained increasing popularity. By analyzing variation channel information (CSI) Wi-Fi signals, states a target person under coverage, such as cough, sneeze, yawn, can be accurately detected. A major problem current is being overly domain-dependent. That is,...

Person re-identification (ReID) is a challenging task that aims to identify individuals across multiple non-overlapping camera views. To enhance the performance and robustness of ReID models, it crucial train them over data sources. However, traditional centralized approach poses significant challenge privacy as requires collecting from distributed owners. overcome this challenge, we employ federated learning approach, which enables model training without compromising privacy. In paper,...

Motivated by the success of deep learning, semantic communication has emerged as new paradigm shifts in 6G from conventional data-oriented communications. However, systems suffer performance degradation at receiver side or computation latency accumulation transmitter side, when serves multiple task execution. To address issue, we develop a VisionTransformer based multi-device system called MDSC to effectively perform tasks. In particular, introduce shared encoder extract global information,...

The rapid adoption of electric vehicles (EVs) stimulates the proliferation charging stations (CSs), motivating cooperative management growing CSs. However, CS still remains an open problem, due to uncertain user behavior and heterogeneous service capabilities. To capture dynamics caused by behavior, we propose a deep reinforcement learning (DRL)-based method for multiple CSs, towards maximizing total profit. proposed determines pricing scheduling decisions considering stochastic CSs...

Semantic communication has emerged as new paradigm shifts in 6G from the conventional syntax-oriented communications. Recently, wireless broadcast technology been introduced to support semantic system toward higher efficiency. Nevertheless, existing systems target on general representation within one stage and fail balance inference accuracy among users. In this paper, encoding process is decomposed into compression fusion improves efficiency with adaptation tasks channels.Particularly, we...

This paper considers a device-edge co-inference system for image classification tasks. The splits deep learning model between an edge device (ED) and server (ES), where the ED feeds raw images to its local transmits output semantic features ES classification. We address two key challenges in transmitting continuous through modern digital wireless communication systems. First, we devise smooth sine-based surrogate function approximate non-differentiable quantization computation applied...