Edge Intelligence in Smart Grids: A Survey on Architectures, Offloading Models, Cyber Security Measures, and Challenges

Abstract

The rapid development of new information and communication technologies (ICTs) and the deployment of advanced Internet of Things (IoT)-based devices has led to the study and implementation of edge computing technologies in smart grid (SG) systems. In addition, substantial work has been expended in the literature to incorporate artificial intelligence (AI) techniques into edge computing, resulting in the promising concept of edge intelligence (EI). Consequently, in this article, we provide an overview of the current state-of-the-art in terms of EI-based SG adoption from a range of angles, including architectures, computation offloading, and cybersecurity concerns. The basic objectives of this article are fourfold. To begin, we discuss EI and SGs separately. Then we highlight contemporary concepts closely related to edge computing, fundamental characteristics, and essential enabling technologies from an EI perspective. Additionally, we discuss how the use of AI has aided in optimizing the performance of edge computing. We have emphasized the important enabling technologies and applications of SGs from the perspective of EI-based SGs. Second, we explore both general edge computing and architectures based on EI from the perspective of SGs. Thirdly, two basic questions about computation offloading are discussed: what is computation offloading and why do we need it? Additionally, we divided the primary articles into two categories based on the number of users included in the model, either a single user or a multiple user instance. Finally, we review the cybersecurity threats with edge computing and the methods used to mitigate them in SGs. Therefore, this survey comes to the conclusion that most of the viable architectures for EI in smart grids often consist of three layers: device, edge, and cloud. In addition, it is crucial that computation offloading techniques must be framed as optimization problems and addressed effectively in order to increase system performance. This article typically intends to serve as a primer for emerging and interested scholars concerned with the study of EI in SGs.

Edge Offloading in Smart Grid

“…The choice between the two frameworks depends on factors such as the specific requirements of the project, the level of detail needed, and the area of interest, as well as on their advantages and disadvantages (see Table 1). However, more specific architectures have been designed to address different challenges related to the computational continuum in more depth, such as data management distribution latency and security [2,3,41]. It is worth noting that even though those architectures have benefits related to edge offloading, the implementation can be a complex process and might need significant upfront investment for the initial setup [42,43].…”

“…Molokomme et al proposed an architecture involving multiple components such as residential, commercial, and industrial devices, edge servers, power systems, IoT devices, and the overall cloud infrastructure [3]. The architecture integrates with edge computing, introducing intelligence for analysis, monitoring, and processing of data at the network's edge.…”

“…However, in smart-grid decentralized scenarios, the offloading of processing workloads towards the edge nodes is challenging due to the heterogeneity, diversity of resources, application characteristics, and edge uncertainty [1]. Challenges like real-time data-processing, reducing latency, and security need to be systematically addressed in the smart grid, and edge and fog computing can play a fundamental role in energy sector decentralization [3].…”

Enabling Technologies for Next-Generation Smart Cities: A Comprehensive Review and Research Directions

“…Sensors, IoT gadgets, and network connections are the foundation of smart homes Smart grid: This comprises three main systems, namely, generation, transportation, and consumption [18]. The consumer system deploys sensing devices that control and monitor the machines on premises.…”

Towards Risk-Free Trustworthy Artificial Intelligence: Significance and Requirements

“…Focussing on improving the conventional Hoepman's eight privacy design strategies, a previous study [218] demonstrated the successful application of such a privacy-preserving mechanism for private and secure IoT services, which included the preservation of data privacy (data collection, storage, and application), user authentication, communication, and computation. Tis simple yet efective design benefts numerous applications, such as identifcation systems, access control systems, smart safety systems, smart grids, and healthcare systems [231].…”