Secure Data Transmission in Smart Meters Using Q-Learning in Fog Computing Environment

With the advancement in technology and connectivity, the concept of smart grids with smart meters has significantly increased the interest of both electric consumers and power suppliers. Nowadays the smart cities are using the technology of smart meters in their household. Which can transfer the electrical data in a bi-directional way to consumers and smart grids. It consists of various internet-connected sensors, software applications, to connect different devices. This makes the smart grid network more complex which is further vulnerable to the cyber-attack. Moreover, it gives rise to various malicious attack such as false data injection (FDI) in smart meters which leads to incorrect decisions due to a large number of open access points in a network which is accessible to outside intruders. In smart meters, Home Area Network (HAN) and smart meter collectors are more vulnerable for cyber-attacks as they are directly involved in bi-directional communication between the devices such as smart grid, data connectors and smart meters. Since smart meters work in a distributed environment where fog computing (FC) placed at the edge of the networks can play a major role by integrating with blockchain for secure electric data transmission and to avoid the FDI in smart meters. Here FC nodes will act as miners to makes the decisions on the data transmission while monitoring the states of the system. Hence, we propose a novel solution for the above mention issue which consists of 3-tier FC-based blockchain architecture and Smart Meter-based Q-Learning Encryption (SMQE) algorithm to verify the system in FC environment. The algorithm performs the identification of FDI, and malicious attacks conducted on smart meters during data transmission. It opens a blockchain-based channel for secure data transmission in smart meters. The simulations tools used for this purpose are iFogSim, SimBlock, Python (Anaconda) with Geth version 1.8.23.