This paper presents an error performance analysis and a model of a narrow-band power line carrier (PLC) system for smart metering. Our work is founded on complex analysis based on the probability theory using limited, long-term measurement data of a rural 400 V distribution grid during operation. To obtain confident results, the analysis and modeling of the error performance were done in two steps. In the first step, the Neyman contagious distribution, originally derived in the fields of entomology and bacteriology, was applied to describe the probability distribution of errors in messages in consideration of the impulsive noise in the PLC channel and the influence of forward error correction techniques. In the second step, assuming the bit error rate (BER) was a random variable, where errors are randomly distributed in the sample rather than clustered into messages, the confidence interval of the true BER was calculated for different SNR values. The results served as a foundation for the error performance model proposed in this paper. The presented work is crucial for the research of upper layer communication protocols performance incorporating advanced phenomena at the physical layer.
Read full abstract