IoT-ReliableComm: A Self-Supervised Approach to Signal Transmission Reliability in Interconnected Consumer Electronics

Abstract

In the Internet of Things (IoT), the proliferation of smart, interconnected consumer electronics (CE) has heightened the demand for reliable signal transmission. However, noise interference continues to be a critical challenge that can impact the overall performance of IoT-enabled devices and the stability of communication. This study introduces sub-sampling denoising compensation network (SDCN) engineered to enhance signal reliability within the IoT ecosystem. SDCN employs a self-supervised learning approach, eliminating the need for traditional paired training datasets. It incorporates a sub-sampler to create training signal pairs, a denoising network for the noise reduction, and a signal residual compensation module to preserve the original signal’s characteristics. This comprehensive solution ensures that signals transmitted between devices remain complete and accurate, even in the presence of substantial noise. The framework’s effectiveness is validated through a series of experiments, demonstrating its superiority in terms of signal transmission reliability.