Lightweight Cryptographic Algorithms for Resource-Constrained IoT Devices: A Security and Performance Evaluation

The rapid expansion of Internet of Things (IoT) systems has introduced significant security challenges due to the constrained computational, memory, and energy resources of embedded devices. Conventional cryptographic algorithms are often unsuitable for such environments, necessitating the use of lightweight cryptographic primitives optimized for efficiency. This paper presents a systematic security and performance evaluation of selected lightweight cryptographic algorithms under an embedded execution environment. Specifically, SIMON-64/128, PRESENT-80, and ASCON-128 are implemented in C and evaluated on an ESP32 microcontroller based on the Xtensa LX6 architecture using the ESP-IDF framework. Performance evaluation is conducted using cycle-accurate hardware measurements, including execution time, CPU cycles, cycles-per-byte (CPB), and throughput under different compiler optimization levels (-Og and -O2). In addition, security properties are analyzed using diffusion-based metrics such as avalanche effect and key sensitivity. Experimental results demonstrate that SIMON achieves the highest computational efficiency, while PRESENT exhibits an unexpected performance regression under higher compiler optimization, highlighting the impact of architecture-compiler interactions. The findings emphasize that algorithm suitability in IoT systems depends not only on theoretical design but also on implementation-level behavior under specific hardware and compiler configurations. This work provides a reproducible evaluation framework and offers practical insights for selecting and optimizing lightweight cryptographic algorithms in resource-constrained embedded systems, forming a foundation for future research toward the design of more efficient lightweight encryption schemes.