The development of Industrial Internet Things (IIoT) and Industry 4.0 has completely changed the traditional manufacturing industry. Intelligent IIoT technology usually involves a large number intensive computing tasks. Resource-constrained devices often cannot meet real-time requirements these As promising paradigm, mobile-edge (MEC) system migrates computation tasks from resource-constrained to nearby MEC servers, thereby obtaining lower delay energy consumption. However, considering...

With the development of technologies deployed on vehicles, there is a significant increase in amount data, which comes from various applications, such as battery management, VR, autopilot, etc. However, privacy critical obstacle to utilizing information since many vehicle-based applications involve locations, conversations, driving behaviors, Federated Learning (FL) promising technology perfect for filling gap, it keeps users' data their devices, gives rise Vehicular Networks (FVN). As...

The energy harvesting cognitive wireless sensor network (EHCWSN) introduces technology and radio into the traditional (WSN), which significantly prolongs working life of node effectively alleviates congestion problem unlicensed spectrum. Due to uncertainty process behavior primary user (PU), how allocate manage limited resources is a crucial in EHCWSN. In this work, new Q-learning-based channel selection method proposed for randomness PU's network. By continuously interacting learning with...

As the segment of diseased tissue in PET images is time-consuming, laborious and low accuracy, this work proposes an automated framework for image screening, denoising segmentation. First, taking into account characteristics images, uses a differential activation filter to select whole-body containing lesion tissue. Second, new neural network residual connections which has powerful generalization performance compared with normal FCN proposed reconstruction denoising. Finally, segmentation...

Telematics technology development offers vehicles a range of intelligent and convenient functions, including navigation mapping services, driving assistance, traffic management. However, since these functions deal with sensitive information like vehicle location habits, it is crucial to address concerns regarding security privacy protection. Federated learning (FL) highly suitable for addressing such problems due its characteristics, in which client does not need share private data upload...

The online social network in wireless communication has become increasingly important owing to the fact that approximately billions of people send and receive information using mobile devices. Communication may face a challenge arising from conflict between traditional theories new research methods data transmission. massive amount transmitted results energy over-consumption, low delivery ratio, transmission delays. This study attempts solve these drawbacks by proposing method effectively...

Unmanned-aerial-vehicle (UAV)-aided target tracking has been applied in many important practical scenarios such as vehicle missions. However, the limited computation capability of UAVs can hardly support computation-intensive tasks, like with real-time video processing. Inspired by strong capabilities edge computing servers nowadays, this article develops an energy-efficient UAV-aided system, where processing tasks be offloaded from a UAV to nodes (ENs) along its flight trajectory. To select...

Serverless computing adopts a pay-as-you-go billing model where applications are executed in stateless and shortlived containers triggered by events, resulting reduction of monetary costs resource utilization. However, existing platforms do not provide an upper bound for the which makes overall cost unpredictable, precluding many organizations from managing their budgets. Due to diverse ranges serverless functions heterogeneous capacity edge devices, it is challenging receive near-optimal...

In social opportunistic networks, mobile devices can be regarded as socialization nodes. Furthermore, they carry and store useful information. Mobile select destination nodes deliver messages through networks because securely conveniently stored, carried, transmitted with According to such characteristics, communities established. However, many may only depending on one or two If those are not enough cache over-flooding, data transmission in wait for a long time. To find cooperation neighbor...

There are many factors affecting the survival of people in developing countries, such as tremendous number population, nonuniform medical resources, and threatening malignant diseases. The improvements information system countries may lead to a bright future. By using effect resources utilizing coming from system, doctors could come diagnosis with analysis. probability getting sick is very useful which assists improve accuracy disease diagnosis, shortening treatment time, reducing incidence...

Edge computing is an emerging novel computation paradigm, which plays a vital role in the 5G era. A basic application scenario edge that end users have ability to choose server nodes, are expected participate working on tasks allocated by users. At same time, nodes can freely make their own decisions serving, user's task. End compete with each other for nodes' service, through payment they offer service. In this paper, we apply game theory model interaction between and as noncooperative...

Along with the coming of Internet Everything era, massive numbers pervasive connected devices in various fields are continuously producing oceans time series stream data. In order to carry out different kinds data mining tasks (similarity search, classification, clustering, and prediction) based on streaming efficiently effectively, segmentation representation which segment a into several subsequences provide approximative for raw data, should be done as first step. With virtue solid...

The hyperparameters of the metaheuristic algorithm are difficult to determine when solving optimization problems. existing methods mainly adjust through preset rules or traditional RL. performance above is unsatisfactory and generalization poor. This work proposes a deep Q-learning network (DQN)-based dynamic setting framework for combinatorial hyperparameters, applies it Genetic (GA) improve its performance. By defining four elements environment, state, action reward required learning...

Handover is a crucial issue in ensuring the continuity of edge services computing (EC) systems. Failure to handle hand-off properly may result delays, data loss, or service interruption during switching. Therefore, optimizing process ensure and satisfactory user experience significant challenge design In this paper, we propose resource reservation algorithm that reserves portion resources each base station meet quality (QoS) requirements We use an LSTM model predict number new existing users...

Target tracking is an important issue of social security. In order to track a target, traditionally large amount surveillance video data need be uploaded into the cloud for processing and analysis, which put stremendous bandwidth pressure on communication links in access networks core networks. At same time, long delay wide area network very likely cause system lose its target. Often, unmanned aerial vehicle (UAV) has been adopted target due flexibility, but limited flight time battery...

This paper proposes an energy management strategy that can resist DOS attacks for solving the Economic Dispatch Problem (EDP) of smart grid. We use concept agent, which acts as a hub grid, and each EA is integrated unit converts, stores, utilizes its local resources. approach takes into account coupling relationship between agents (EA) Lyapunov function technique to achieve finite-time solutions optimization problems. incorporate strategies when analyzing convergence using technique. Based...

Adopting serverless computing to edge networks benefits end-users from the pay-as-you-use billing model and flexible scaling of applications. This paradigm extends boundaries remarkably improves quality services. However, due heterogeneous nature bandwidth resources in networks, it is challenging dynamically allocate different while adapting burstiness high concurrency workloads. article focuses on function provisioning optimize end-to-end latency, where challenge lies jointly allocating...

Variational Autoencoders (VAEs), as a form of deep generative model, have been widely used in recent years, and shown great peformance number different domains, including image generation anomaly detection, etc.. This paper aims to explore neural network model compression method based on VAE. The experiment uses models for MNIST recognition targets, Feedforward Neural Network (FNN), Convolutional (CNN), Recurrent (RNN) Long Short-Term Memory (LSTM). These are the most basic learning, other...