A quantum-resistant advanced metering infrastructure

Abstract

This dissertation focuses on discussing and implementing a Quantum-Resistant Advanced Metering Infrastructure (QR-AMI) that employs quantum-resistant asymmetric and symmetric cryptographic schemes to withstand attacks from both quantum and classical computers. The proposed solution involves the integration of Quantum-Resistant Dedicated Cryptographic Modules (QR-DCMs) within Smart Meters (SMs). These QR-DCMs are designed to embed quantum-resistant cryptographic schemes suitable for AMI applications. In this sense, it investigates quantum-resistant asymmetric cryptographic schemes based on strong cryptographic principles and a lightweight approach for AMIs. In addition, it examines the practical deployment of quantum-resistant schemes in QR-AMIs. Two candidates from the National Institute of Standards and Technology (NIST) post-quantum cryptography (PQC) standardization process, FrodoKEM and CRYSTALS-Kyber, are assessed due to their adherence to strong cryptographic principles and lightweight approach. The feasibility of embedding these schemes within QRDCMs in an AMI context is evaluated through software implementations on low-cost hardware, such as microcontroller and processor, and hardware/software co-design implementations using System-on-a-Chip (SoC) devices with Field-Programmable Gate Array (FPGA) components. Experimental results show that the execution time for FrodoKEM and CRYSTALS-Kyber schemes on SoC FPGA devices is at least one-third faster than software implementations. Furthermore, the achieved execution time and resource usage demonstrate the viability of these schemes for AMI applications. The CRYSTALS-Kyber scheme appears to be a superior choice in all scenarios, except when strong cryptographic primitives are necessitated, at least theoretically. Due to the lack of off-the-shelf SMs supporting quantum-resistant asymmetric cryptographic schemes, a QRDCM embedding quantum-resistant scheme is implemented and evaluated. Regarding hardware selection for QR-DCMs, microcontrollers are preferable in situations requiring reduced processing power, while SoC FPGA devices are better suited for those demanding high processing power. The resource usage and execution time outcomes demonstrate the feasibility of implementing AMI based on QR-DCMs (i.e., QR-AMI) using microcontrollers or SoC FPGA devices.