Joint Service Caching and Computation Offloading Scheme Based on Deep Reinforcement Learning in Vehicular Edge Computing Systems

IEEE Transactions on Vehicular Technology, Journal Year: 2023, Volume and Issue: 72(5), P. 6709 - 6722

Published: Jan. 5, 2023

Vehicular edge computing (VEC) is a new paradigm that enhances vehicular performance by introducing both computation offloading and service caching, to resource-constrained vehicles ubiquitous servers. Recent developments of autonomous enable variety applications demand high resources low latency, such as automatic driving, auto navigation, etc. However, the highly dynamic topology networks limited caching space at servers calls for intelligent design placement offloading. Meanwhile, decisions are correlated decisions, which pose great challenge effectively strategies. In this paper, we investigate joint optimization problem integrating in general VEC scenario with time-varying task requests. To minimize average processing delay, formulate using long-term mixed integer non-linear programming (MINLP) propose an algorithm based on deep reinforcement learning obtain suboptimal solution complexity. The simulation results demonstrate our proposed scheme exhibits effective improvement delay compared other representative benchmark methods.

Language: Английский

---

Edge Intelligence in Intelligent Transportation Systems: A Survey

IEEE Transactions on Intelligent Transportation Systems, Journal Year: 2023, Volume and Issue: 24(9), P. 8919 - 8944

Published: May 24, 2023

Edge intelligence (EI) is becoming one of the research hotspots among researchers, which believed to help empower intelligent transportation systems (ITS). ITS generates a large amount data at network edge by millions devices and sensors. Data-driven artificial (AI) core development. By pushing AI frontier edge, EI enables applications have lower latency, higher security, less pressure on backbone better use big data. This paper surveys Intelligence in Intelligent Transportation Systems. We first introduce challenges faces explain motivation using ITS. then explore framework ITS, including EI-based architecture, gathering communication methods, processing service delivery, performance indexes. The enabling technologies, such as models, Internet Things, Computing technologies used are reviewed intensively. discuss fields depth. Typical application scenarios, autonomous driving, vehicular computing, system, unmanned aerial vehicle (UAV) environment, rail control management, explored. general platforms EI, training inference well benchmark datasets, introduced. Finally, we some future directions

Language: Английский

---

Reinforcement Learning Methods for Computation Offloading: A Systematic Review

ACM Computing Surveys, Journal Year: 2023, Volume and Issue: 56(1), P. 1 - 41

Published: June 9, 2023

Today, cloud computation offloading may not be an appropriate solution for delay-sensitive applications due to the long distance between end-devices and remote datacenters. In addition, a can consume bandwidth dramatically increase costs. However, such as sensors, cameras, smartphones have limited computing storage capacity. Processing tasks on battery-powered energy-constrained devices becomes even more complex. To address these challenges, new paradigm called Edge Computing (EC) emerged nearly decade ago bring resources closer end-devices. Here, edge servers located end-device perform user tasks. Recently, several paradigms Mobile (MEC) Fog (FC) complement Cloud (CC) EC. Although are heterogeneous, they further reduce energy consumption task response time, especially applications. Computation is multi-objective, NP-hard optimization problem. A significant part of previous research in this field devoted Machine Learning (ML) methods. One essential types ML Reinforcement (RL), which agent learns how make best decision using experiences gained from environment. This article provides systematic review widely used RL approaches offloading. It covers complementary mobile computing, fog Internet Things. We explain reasons various methods technical point view. analysis includes both binary partial techniques. For each method, elements characteristics environment discussed regarding most important criteria. Research challenges Future trends also mentioned.

Language: Английский

---

Federated Learning for Computational Offloading and Resource Management of Vehicular Edge Computing in 6G-V2X Network

IEEE Transactions on Consumer Electronics, Journal Year: 2024, Volume and Issue: 70(1), P. 3827 - 3847

Published: Jan. 26, 2024

The Sixth Generation network (6G) can support autonomous driving along with various vehicular applications like Vehicular Edge Computing (VEC), a distributed computing architecture for connected vehicles. Computational offloading and resource management of help sort out some issues, such as high communication costs, privacy protection, an excessively long training process, etc., by proposing efficient model the Federated Learning computational in environment. Two research issues are highlighted this paper. One problem is related to current system: smart structure operating system. Consistent access cloud services, regardless installed system or used hardware, still challenging. Another issue security privacy. Security two important features that should be maintained data centers transmission during management. In survey paper, going proposed which will give partial solution these issues. solution, found while conducting review, offers train update edge devices' information. entire provide updated information devices solve difficulties getting key necessary model-related optimization. This also enhance effectiveness frameworks 6G-V2X communication.

Language: Английский

---

Integration of data science with the intelligent IoT (IIoT): current challenges and future perspectives

Digital Communications and Networks, Journal Year: 2024, Volume and Issue: unknown

Published: March 1, 2024

The Intelligent Internet of Things (IIoT) involves real-world things that communicate or interact with each other through networking technologies by collecting data from these "things" and using intelligent approaches, such as Artificial Intelligence (AI) machine learning, to make accurate decisions. Data science is the dealing its relationships approaches. Most state-of-the-art focus on topic independently, either IIoT. Therefore, address gap, this article provides a comprehensive survey advances integration IoT system classifying existing IoT-based techniques presenting summary various characteristics. paper analyzes big security privacy features, including network architecture, protection, continuous monitoring data, which face challenges in systems. Extensive insights into security, privacy, are visualized context for IoT. In addition, study reveals current opportunities enhance market development. gap faced comprehensively presented, followed future outlook possible solutions challenges.

Language: Английский

---

Federated Learning for Edge Computing: A Survey

Applied Sciences, Journal Year: 2022, Volume and Issue: 12(18), P. 9124 - 9124

Published: Sept. 11, 2022

New technologies bring opportunities to deploy AI and machine learning the edge of network, allowing devices train simple models that can then be deployed in practice. Federated (FL) is a distributed technique create global model by from multiple decentralized clients. Although FL methods offer several advantages, including scalability data privacy, they also introduce some risks drawbacks terms computational complexity case heterogeneous devices. Internet Things (IoT) may have limited computing resources, poorer connection quality, or use different operating systems. This paper provides an overview used with focus on resources. presents frameworks are currently popular provide communication between clients servers. In this context, various topics described, which include contributions trends literature. includes basic designs system architecture, possibilities application practice, privacy security, resource management. Challenges related requirements such as hardware heterogeneity, overload resources discussed.

Language: Английский

---

A Novel Hybrid Deep Learning Model for Metastatic Cancer Detection

Computational Intelligence and Neuroscience, Journal Year: 2022, Volume and Issue: 2022, P. 1 - 14

Published: June 24, 2022

Cancer has been found as a heterogeneous disease with various subtypes and aims to destroy the body’s normal cells abruptly. As result, it is essential detect prognosis distinct type of cancer since they may help survivors treatment in early stage. It must also divide patients into high- low-risk groups. While realizing efficient detection frequently time-taking exhausting task high possibility pathologist errors previous studies employed data mining machine learning (ML) techniques identify cancer, these strategies rely on handcrafted feature extraction that result incorrect classification. On contrary, deep (DL) robust recently widely used for classification purposes. This research implemented novel hybrid AlexNet-gated recurrent unit (AlexNet-GRU) model lymph node (LN) breast We have well-known Kaggle (PCam) set classify LN samples. study tested compared among three models: convolutional neural network GRU (CNN-GRU), CNN long short-term memory (CNN-LSTM), proposed AlexNet-GRU. The experimental results indicated performance metrics accuracy, precision, sensitivity, specificity (99.50%, 98.10%, 98.90%, 97.50) can reduce occur during diagnosis process significantly better than CNN-GRU CNN-LSTM models. other recent ML/DL algorithms analyze model’s efficiency, which reveals AlexNet-GRU computationally efficient. Also, presents its superiority over state-of-the-art methods

Language: Английский

---

A Learning-Based Approach for Vehicle-to-Vehicle Computation Offloading

IEEE Internet of Things Journal, Journal Year: 2022, Volume and Issue: 10(8), P. 7244 - 7258

Published: Dec. 13, 2022

Vehicle-to-vehicle (V2V) computation offloading has emerged as a promising solution to facilitate computing-intensive vehicular task processing, where vehicles (i.e., TaVs) will be requested offload tasks server SeVs) in order keep delay low. However, it is challenging for TaVs obtain the optimal V2V decisions realizing minimal delay) due constraints, including: 1) incomplete information; 2) degraded Quality-of-Service (QoS) of SeVs; and 3) privacy leakage risks. In this article, we develop learning-based algorithm enhanced by SeV's ability & trustfulness awareness solve these problems. We emphasize that proposed learns performance candidate SeVs based on history selections, without requiring complete information advance. Additionally, both QoS safe are algorithm. Furthermore, conduct extensive simulation experiments validate The results demonstrate reduces average 35% 40%, at same time decreases learning regret 39% 41%, compared algorithms awareness.

Language: Английский

---

A survey on vehicular task offloading: Classification, issues, and challenges

Journal of King Saud University - Computer and Information Sciences, Journal Year: 2022, Volume and Issue: 34(7), P. 4135 - 4162

Published: May 25, 2022

Emerging vehicular applications with strict latency and reliability requirements pose high computing requirements, current vehicles' computational resources are not adequate to meet these demands. In this scenario, vehicles can get help process tasks from other resource-rich platforms, including nearby vehicles, fixed edge servers, remote cloud servers. Nonetheless, different communication network (VCN) modes need be utilized access resources, improving networks' performance quality of service (QoS). paper, we present a comprehensive survey on the task offloading techniques under perspective, i.e., vehicle (V2V), roadside infrastructure (V2I), everything (V2X). For task/computation offloading, classification methods V2V, V2I, V2X domains. Besides, categories each sub-categorized according their schemes' objectives. Furthermore, literature is elaborated, compared, analyzed perspectives approaches, objectives, merits, demerits, etc. Finally, highlight open research challenges in field predict future trends.

Language: Английский

---

Towards Risk-Free Trustworthy Artificial Intelligence: Significance and Requirements

International Journal of Intelligent Systems, Journal Year: 2023, Volume and Issue: 2023, P. 1 - 41

Published: Oct. 26, 2023

Given the tremendous potential and influence of artificial intelligence (AI) algorithmic decision-making (DM), these systems have found wide-ranging applications across diverse fields, including education, business, healthcare industries, government, justice sectors. While AI DM offer significant benefits, they also carry risk unfavourable outcomes for users society. As a result, ensuring safety, reliability, trustworthiness becomes crucial. This article aims to provide comprehensive review synergy between DM, focussing on importance trustworthiness. The addresses following four key questions, guiding readers towards deeper understanding this topic: (i) why do we need trustworthy AI? (ii) what are requirements In line with second question, that establish been explained, explainability, accountability, robustness, fairness, acceptance AI, privacy, accuracy, reproducibility, human agency, oversight. (iii) how can data? (iv) priorities in terms challenging applications? Regarding last six different discussed, environmental science, 5G-based IoT networks, robotics architecture, engineering construction, financial technology, healthcare. emphasises address before their deployment order achieve goal good. An example is provided demonstrates be employed eliminate bias resources management systems. insights recommendations presented paper will serve as valuable guide researchers seeking applications.

Language: Английский

---