A Unified Approach for Compression and Authentication of Smart Meter Reading in AMI

Abstract

In this paper, we propose a unified approach for compression and authentication of smart meter reading in advanced metering infrastructure (AMI). In general, smart meters are urged to send sampled reading signals at a high rate for high-quality services. Meanwhile, power reading signals have to be authenticated to prevent impersonation attacks, which can cause serious economic loss. However, the security in smart grids faces more challenges than conventional human-type communications because of limited hardware resources of a smart meter (e.g., small memory). Motivated by these problems, we study simultaneous compression and authentication for power reading signals in multicarrier systems based on the notion of compressive sensing (CS). The CS-based compression and authentication method are applied to empirically modeled signals with a shared secret key, a measurement matrix in CS between a data concentrator unit (DCU) and a legitimate smart meter. In particular, for authentication, the residual error of a received signal at the DCU is used as a test statistic for hypothesis testing, which determines whether the signal is a legitimate signal or an intrusion signal in the proposed approach. Through the analysis and simulation results, we demonstrate that the CS-based compression approach can be applied to smart meter reading with good energy efficiency. In addition, it is shown that the proposed scheme can obtain a low authentication error probability under reasonable conditions. For example, when the number of subcarriers is 64, the DCU can distinguish legitimate and intrusion smart meters with a probability of $1 - P_{E}$ , where $P_{E} \leq 10^{-4}$ .