Abstract
Power line communication (PLC) systems are essential for modern telecommunications, providing a cost-effective solution for data transmission over existing electrical wiring. However, indoor environments pose significant challenges for these systems, primarily due to burst errors caused by impulsive noise, narrowband interference, and signal attenuation. This study seeks to address these challenges by utilizing three-state Semi-Hidden Markov Models (SHMMs) to analyze and characterize burst error processes in narrowband PLC (NB-PLC) systems. We developed a software-defined multi-state quadrature amplitude modulation (M-QAM) orthogonal frequency division multiplexing (OFDM)-based NB-PLC testbed that adheres to the International Telecommunication Union - Telecommunication Standardization Sector (ITU-T) G.hnem standard for home networking over existing wiring, such as power lines, phone lines, and coaxial cables. The system incorporates 4-QAM-OFDM, 8-QAM-OFDM, and 16-QAM-OFDM modulation schemes, utilizing Universal Software Radio Peripheral (USRP) hardware for real-time transmission and reception. Using this experimental setup, we generated and analyzed a dataset of 200,000 transmitted bits, capturing the characteristics of burst errors under realistic indoor noise conditions. A comparative analysis of reference error sequences and SHMM simulations demonstrates the model's ability to accurately replicate NB-PLC-specific error dynamics, including error clustering and state transitions. This research makes four key contributions: (1) the development of a robust experimental testbed for real-time burst error analysis in NB-PLC systems, (2) the creation of a comprehensive dataset that characterizes noise-induced disruptions in real-world indoor scenarios, (3) a theoretical exploration and practical application of SHMMs for modeling error sequences in NB-PLC channels, and (4) the validation of SHMMs and a modified Baum-Welch algorithm (BWA) for efficient modeling and parameter estimation of burst error processes. Together, these contributions enhance the understanding of burst error dynamics and provide a foundation for improving the resilience and efficiency of NB-PLC networks in real-world indoor environments.