Blockchain-Enabled Intelligent Vehicular Edge Computing

【Author】 Islam, Shafkat; Badsha, Shahriar; Sengupta, Shamik; La, Hung; Khalil, Ibrahim; Atiquzzaman, Mohammed

【Source】IEEE NETWORK

【影响因子】10.294

【Abstract】Smart vehicles are expected to be equipped with high-dimensional, resource-intensive applications, including platoon control, augmented reality supported gaming, AI-based pedestrian detection, fuel scheduling, and so on, catering to diverse user preferences and enhancing safety and efficiency. These applications pose unique challenges for resource-constrained vehicles due to their intense computation requirements, whereas vehicular edge computing (VEC) networks, consisting of roadside units (RSUs) and MEC servers, contain the capability of providing cloud-like computing experience at vehicular edges while meeting performance requirements in terms of latency and throughput. Moreover, the development of intelligent VEC (IVEC) infrastructure is accelerated due to rapid advancement of AI algorithms in recent years. However, IVEC is prone to attacks, including fake computation feedback, unfair or biased resource allocation in a VEC server, and so on, due to its centralized governance and black box computation (edge computation works like a black box for end users). To combat such security vulnerabilities, we propose a blockchain-based decentralized architecture to enhance transparency in IVEC resource management and leverage edge consumers (e.g., vehicles) with a computation verification option. Additionally, we address the unbalanced load distribution issue and propose a secure IVEC federation model for balancing loads. We also outline the main challenges and provide a brief description of promising research directions to draw the attention of concerned stakeholders and parties in both the blockchain and edge computing domains.

【Keywords】Computational modeling; Blockchain; Computer architecture; Computational efficiency; Resource management; Servers; Security

【发表时间】2021 MAY-JUN

【收录时间】2022-01-02

【文献类型】

【主题类别】

--