| Abstract |
To address security and privacy challenges in the Internet of Things (IoT), coverless video steganography methods have been explored for hiding sensitive data in sectors like manufacturing, healthcare, agriculture, and smart cities. However, existing methods lack thorough analysis of their suitability for IoT, given the limited energy, connectivity, and hardware capabilities of these devices. This paper presents a performance analysis of a hybrid self-sufficient coverless video steganography (HCVSteg) method, tested with Raspberry Pi 3 Model B nodes over a wireless network using the HMDB51 and UCF101 datasets. The HCVSteg method demonstrated exceptional efficiency, using 97.35% less CPU and less than 2% of memory compared to state-of-the-art lightweight security algorithms while accelerating encoding and decoding by 51% and 60%, respectively. The novel synthesis scheme reduced computational demands related to missing byte sequences. Moreover, the method consumed only 0.01W, making it 37 to 1,677 times more energy-efficient than comparable algorithms. Transmission rate improved by 6%, latency decreased by 6%, and the method added a minimal overhead of 100B for retrieval information. These results emphasize the method’s suitability in resource-constrained environments like IoT, where speed, reliability, and energy efficiency are critical. © 2024 Copyright held by the owner/author(s). |
