Ad Hoc Vehicular Fog Enabling Cooperative Low-Latency Intrusion Detection

Internet of Vehicles and vehicular networks have been compelling targets for malicious security attacks where several intrusion detection solutions have been proposed for protecting them. Nonetheless, their main problem lies in their heavy computation, which makes them unsuitable for next-generation artificial intelligence-powered self-driving vehicles whose computational power needs to be primarily reserved for real-time driving decisions. To address this challenge, several approaches have been lately presented to take advantage of the cloud computing for offloading intrusion detection tasks to central cloud servers, thus reducing storage and processing costs on vehicles. However, centralized cloud computing entails high latency on intrusion detection related data transmission and plays against its adoption in delay-critical intelligent applications. In this context, this article proposes a vehicular-edge computing (VEC) fog-enabled scheme allowing offloading intrusion detection tasks to federated vehicle nodes located within nearby formed ad hoc vehicular fog to be cooperatively executed with minimal latency. The problem has been formulated as a multiobjective optimization model and solved using a genetic algorithm maximizing offloading survivability in the presence of high mobility and minimizing computation execution time and energy consumption. Experiments performed on resource-constrained devices within actual ad hoc fog environment illustrate that our solution significantly reduces the execution time of the detection process while maximizing the offloading survivability under different real-life scenarios.

I. Introduction

Internet of Vehicles (IoV), an essential paradigm of the Internet of Things (IoT) [1] related to intelligent transportation systems (ITS), are networks of interconnected vehicles and roadside units (RSUs) that exchange information to detect, prevent, and manage traffic problems [2]. For example, vehicles can exchange information about road conditions in order to avoid traffic jams. Due to their expansion, several commercial applications (e.g., multimedia and infotainment) have also been integrated into such networks for marketing and business purposes. Unfortunately, vehicular networks have been an appealing target for countless malicious attacks (e.g., impersonation and bogus information dissemination) due to their infrastructureless and distributed nature [3]. Such attacks are likely to result in catastrophic consequences ranging from loss of lives (in case of traffic management applications) to loss of revenue (in case of commercial applications) [4]. This demands setting up strict security measures (i.e., intrusion detection and prevention systems) to protect the vehicular network from being an easy target for attackers. This is however a difficult task in such networks that have special restrictions and needs. The main restriction is related to the massive computation load that is often needed for analyzing big intrusion detection data. For example, previous results showed that running an intrusion detection system (IDS) over a data set of 10-MB size can consume up to 212 J of energy, up to 100 MB of RAM and up to 460 s of CPU on a mobile device, with an execution time of 459 s [5]. On the other hand, with the rapid evolution in ITS and the integration of artificial intelligence (AI) capabilities into today’s smart devices (i.e., autonomous vehicles), the resources that are available on smart vehicles have to be efficiently exploited to ensure the success of such technology and avoid undesirable consequences. In other words, the resources that are available on AI-powered vehicles should be primarily dedicated to making real-time driving decisions (e.g., detecting pedestrians, cyclists, etc.), rather than to security assurance. Therefore, there should be a smart load distribution mechanism to efficiently decide where the intrusion detection tasks should be executed. Motivated by this idea, we design in this work a distributed multilayer offloading approach that takes advantage of the emerging edge computing technology to perform intrusion detection in vehicular communication systems in an efficient and resource-aware manner.