Energy-architecture tuning for ECC-based RFID tags

Abstract

The implementation of Elliptic Curve Cryptography (ECC) on small microcontrollers is challenging. Past research has therefore emphasized performance optimization: pick a target architecture, and minimize the cycle count and footprint of the ECC software. This paper addresses a different aspect of resource-constrained ECC implementation: given the application profile, identify the most suitable architecture parameters. At the highest level, an application profile for ECC-based RFID tags is defined by the required security level, signature generation latency and the available energy/power budget. The target architecture parameters of interest include core-voltage, core-frequency, and/or the need for hardware acceleration. The paper brings two contributions to this complex design space exploration problem. First, we introduce a prototype setup for the precise energy measurement of a microcontroller-based ECC implementation. Second, we present a methodology to derive and optimize the architecture parameters starting from the application requirements. We demonstrate our methodology on a MSP430F5438A microcontroller, and present the energy/architecture design space for 80-bit and 128-bit security-levels, for prime field curves secp160r1 and nistp256.