Enhanced cryptosystem for resource-constrained devices with streamlined randomness analysis

As real-world images are increasingly used in digital systems, image data has become a favorite target for malware and advanced cyber-attacks. This highlights the necessity for lightweight and efficient image encryption methods. Towards this end, we introduce an ultralight image encryption scheme based on Cellular Automata (CA), providing strong security with low computational complexity. The algorithm demonstrates strong resistance to differential attacks and successfully passes stringent randomness evaluations, including the Diehard test suite and the NIST statistical standards, at the specified significance levels. The algorithm's performance, featuring a Feistel structure with 128-bit key and data sizes, was assessed on an XC4VLX25 Field-Programmable Gate Array (FPGA). It achieved a processing speed of 2200 Mbits/sec at an operational frequency of 550 MHz, utilizing only 312 slices. This results in a throughput-to-area efficiency of 7.05 Mbits/sec per slice and a bit-per-slice ratio of 0.410. With the reasonable consumption of resources, the new design provides better throughput, lower latency, and better area utilization, making it ideal for real-time and resource-limited environments. Hardware and software simulations both attest to its viability for security-sensitive applications like healthcare and industrial control systems. These results validate the framework's suitability for safe, high-speed image encryption in embedded systems.