Edge computing

About: Edge computing is a research topic. Over the lifetime, 11657 publications have been published within this topic receiving 148533 citations.

Secure Authentication and Key Management With Blockchain in VANETs

Abstract: Nowadays, with rapid advancements of vehicular telematics and communication techniques, proliferation of vehicular ad hoc networks (VANETs) have been witnessed, which facilitates the construction of promising intelligent transportation system (ITS). Due to inherent wireless communicating features in open environment, secure transmission among numerous VANET entities remains crucial issues. Currently, lots of research efforts have been made, while most of which tend to allocate the universal group key to the verified devices for both vehicle-to-vehicle (V2V) and vehicle-to-RSU (V2R) communications. However, in heterogeneous VANET environment with large numbers of devices in same vehicular group, complicated and variable topologies lead to continuous key updating in every moment, causing interference to regular V2R data exchange, which is not reliable and efficient for resource-constrained VANET environment. Moreover, group membership recording and detecting mechanisms are necessary for real time vehicle revocation and participation, which has not been further studied so far. In this paper, we address the above issues by proposing a secure authentication and key management scheme. In our design, novel VANET system model with edge computing infrastructure is adopted so as to offer adequate computing and storing capacity compared to traditional VANET structure. Note that our certificateless authentication scheme applies the independent session key for each vehicle for interference avoidance. Furthermore, consortium blockchain is employed for V2V group key construction. Real time group membership arrangement with efficient group key updating is accordingly provided. Formal security proofs are presented, demonstrating that the proposed scheme can achieve desired security properties. Performance analysis is conducted as well, proving that the proposed scheme is efficient compared with the state-of-the-arts.

Cognitive Radio Network and Network Service Chaining toward 5G: Challenges and Requirements

Abstract: Cognitive radio is a promising technology that answers the spectrum scarcity problem arising from the growth of usage of wireless networks and mobile services. Cognitive radio network edge computing will enhance the CRN capabilities and, along with some adjustments in its operation, will be a key technology for 5G heterogeneous network deployment. This article presents current requirements and challenges in CRN, and a review of the limited research work on the CRN cloud, which will take off CRN capabilities and 5G network requirements and challenges. The article proposes a cognitive radio edge computing access server deployment for network service chaining at the access layer level.

Efficient Federated Learning Algorithm for Resource Allocation in Wireless IoT Networks

Abstract: Federated learning (FL) allows multiple edge computing nodes to jointly build a shared learning model without having to transfer their raw data to a centralized server, thus reducing communication overhead. However, FL still faces a number of challenges such as nonindependent and identically distributed data and heterogeneity of user equipments (UEs). Enabling a large number of UEs to join the training process in every round raises a potential issue of the heavy global communication burden. To address these issues, we generalize the current state-of-the-art federated averaging (FedAvg) by adding a weight-based proximal term to the local loss function. The proposed FL algorithm runs stochastic gradient descent in parallel on a sampled subset of the total UEs with replacement during each global round. We provide a convergence upper bound characterizing the tradeoff between convergence rate and global rounds, showing that a small number of active UEs per round still guarantees convergence. Next, we employ the proposed FL algorithm in wireless Internet-of-Things (IoT) networks to minimize either total energy consumption or completion time of FL, where a simple yet efficient path-following algorithm is developed for its solutions. Finally, numerical results on unbalanced data sets are provided to demonstrate the performance improvement and robustness on the convergence rate of the proposed FL algorithm over FedAvg. They also reveal that the proposed algorithm requires much less training time and energy consumption than the FL algorithm with full user participation. These observations advocate the proposed FL algorithm for a paradigm shift in bandwidth-constrained learning wireless IoT networks.

Design, Resource Management, and Evaluation of Fog Computing Systems: A Survey

Abstract: A steady increase in Internet-of-Things (IoT) applications needing large-scale computation and long-term storage has lead to an overreliance on cloud computing. The resulting network congestion in the cloud, coupled with the distance of cloud data centers from IoT, contributes to unreliable end-to-end response delay. Fog computing has been introduced as an alternative to cloud, providing low-latency service by bringing processing and storage resources to the network edge. In this survey, we sequentially present the phases required in the implementation and realization of practical fog computing systems: 1) design and dimensioning of a fog infrastructure; 2) fog resource provisioning for IoT application use and IoT resource allocation to fog; 3) installation of fog frameworks for fog resource management; and 4) evaluation of fog infrastructure through simulation and emulation. Our focus is on determining the implementation aspects required to build a practical large-scale fog computing infrastructure to support the general IoT landscape.

Machine Learning Technologies for Secure Vehicular Communication in Internet of Vehicles: Recent Advances and Applications

Abstract: Recently, interest in Internet of Vehicles’ (IoV) technologies has significantly emerged due to the substantial development in the smart automobile industries. Internet of Vehicles’ technology enables vehicles to communicate with public networks and interact with the surrounding environment. It also allows vehicles to exchange and collect information about other vehicles and roads. IoV is introduced to enhance road users’ experience by reducing road congestion, improving traffic management, and ensuring the road safety. The promised applications of smart vehicles and IoV systems face many challenges, such as big data collection in IoV and distribution to attractive vehicles and humans. Another challenge is achieving fast and efficient communication between many different vehicles and smart devices called Vehicle-to-Everything (V2X). One of the vital questions that the researchers need to address is how to effectively handle the privacy of large groups of data and vehicles in IoV systems. Artificial Intelligence technology offers many smart solutions that may help IoV networks address all these questions and issues. Machine learning (ML) is one of the highest efficient AI tools that have been extensively used to resolve all mentioned problematic issues. For example, ML can be used to avoid road accidents by analyzing the driving behavior and environment by sensing data of the surrounding environment. Machine learning mechanisms are characterized by the time change and are critical to channel modeling in-vehicle network scenarios. This paper aims to provide theoretical foundations for machine learning and the leading models and algorithms to resolve IoV applications’ challenges. This paper has conducted a critical review with analytical modeling for offloading mobile edge-computing decisions based on machine learning and Deep Reinforcement Learning (DRL) approaches for the Internet of Vehicles (IoV). The paper has assumed a Secure IoV edge-computing offloading model with various data processing and traffic flow. The proposed analytical model considers the Markov decision process (MDP) and ML in offloading the decision process of different task flows of the IoV network control cycle. In the paper, we focused on buffer and energy aware in ML-enabled Quality of Experience (QoE) optimization, where many recent related research and methods were analyzed, compared, and discussed. The IoV edge computing and fog-based identity authentication and security mechanism were presented as well. Finally, future directions and potential solutions for secure ML IoV and V2X were highlighted.