Mitigating inter-pixel interference in MIMO-OCC systems with deep learning: addressing out-of-focus blur and very low-resolution effects

Out-of-focus blur and very low-resolution (VLR) are significant challenges for optical camera communication (OCC) systems that utilize multiple-input multiple-output (MIMO) configurations with LED arrays to increase data rates. These challenges cause inter-pixel interference (IPI), which could severely impair system performance and limit the maximum achievable communication link length. This study experimentally investigates the impact of these challenges on MIMO-OCC systems and proposes a mitigation strategy using a deep learning-based Restormer network. The performance of the Restormer network is compared with some other deep learning techniques. The experimental results show that the suggested method successfully addresses both problems: it increases the signal-to-noise ratio (SNR) by approximately 10 dB under out-of-focus blur conditions, and it extends the maximum link length from 14 to 32 m for a SNR of 12 dB, by mitigating the VLR challenge.