SSDS: A Smart Software-Defined Security Mechanism for Vehicle-to-Grid Using Transfer Learning

Nowadays, vehicle-to-grid (V2G) is a very important component for smart cities, which provide a novel energy storage and scheduling approach. However, security threats currently disturb the normal operations of V2G. There are two challenges for the security of V2G. On one hand, existing security schemes for V2G just consider the static security strategy, which cannot deal well with the problem of high dynamics and advanced persistent threat in V2G. On the other hand, existing V2G lacks a unified information modeling approach, which results in the difficulties of security and communications. To address above challenges, this paper proposes a smart software-defined security mechanism, SSDS, for V2G using transfer learning and IEC 61850 standards. First, as next generation networking technology, software-defined networking (SDN) is adopted to establish a dynamic security protection architecture for V2G, which can provide dynamic security strategy configuration capability. Second, IEC 61850 is introduced to model SSDS, including SDN controller and OpenFlow switch. Third, transfer learning-based security strategy constructing scheme is proposed for the security strategy updating dynamically. Simulation results in terms of network performance and security strategy constructing verify the efficiency as well as feasibility of the proposed security mechanism.