An LLR-Based Receiver for Mitigating Bursty Impulsive Noise With Unknown Distributions

Abstract

The rapid expansion of Internet of Things (IoT) networks has paved the way for their integration into mission-critical applications requiring secure and reliable monitoring, such as smart grid utilities. However, these advanced power grids face significant challenges in maintaining reliable wireless communication, particularly in hostile environments like high-voltage substations and power plants. These environments are characterized by intense bursts of interference, known as impulsive noise with memory. To address this problem, in this study, we introduce a two-process receiver design. The first process is a multi-step receiver parameter estimation process. The second process is a novel memory-aware log-likelihood ratio (LLR) calculation method designed to mitigate the effects of impulsive noise with memory using the parameters estimated from the first process. This method is computationally efficient, which makes it suitable for IoT devices with limited computational capabilities. Simulation results obtained show that the proposed method achieves a bit error rate (BER) similar to the corresponding BERs of the best-performing algorithms with perfect noise parameters. Furthermore, it outperforms the Viterbi algorithm amid imperfect noise parameters. Notably, it method achieves these benchmarks while substantially improving execution time.

Keywords

• Bursty impulsive noise
• log-likelihood ratio
• receiver design
• machine learning
• BCJR algorithm
• Viterbi algorithm