Power line communication (PLC) channel measurements and characterization.

Abstract

The potential of the power line to transport both power and communication signals simultaneously has been realized and practiced for over a century, dating back to the 1900’s. Since the key aspect of power line communications being its expansivity, its implementations were largely as a retrofit technology. This motivation of power line communication is typical for low-, medium-, and high voltage distribution networks. Beyond the “last mile” part, there’s an uprising appeal for intra-building networks currently targeted for home automation (smart homes/buildings) and in-building networking. The optimum use of the existing power line channels has been a focus area for researchers and designers, with the inherent channel hostility proving a serious drawback for high speed data communications. The low-voltage electrical network has unpredictable noise sources, moreover it has two other main disadvantages as a communication channel. The first short coming has to do with the unknown characteristics of the power cable and topology of the network, the second arises from the time-dependent fluctuation of the impedance level of the power line as the loads are switched into and out of the power line network in an unpredictable manner. These factors determine the behaviour of the power line channel when a high frequency signal is impressed on it. This study has shown that the behaviour of indoor power line channels can be captured using a multipath based model even with limited qualitative and/or quantitative knowledge of the network topology. This model is suitable for typical indoor power line channels where knowledge of the topology is near impossible. Some of the feed parameters are obtained through measurements. With sufficient adjustment of control parameters, this model was successfully validated using sample measured channels from the numerous measurements. Through noise measurements, this study has established that impulsive noise is the rifest in the frequency band of interest. The impulsive energy rises well above background noise, which translates to possible data “black outs”. The statistics of the components of this noise are presented. A model of sufficient simplicity is used to facilitate the qualitative description of the background noise through its power spectral density. Two descriptions are provided in terms of the worst and best case scenarios of the background noise occurrences. The model has a good macroscopic capture of the noise power spectral density, with narrow-band interference visible for the worst case noise. Due to the multipath nature of the power line channel, this study also presents the dispersive characteristics of the power line as a communication channel. The power delay profile is used to determine parameters such as first arrival delay, mean excess delay, root mean square delay spread and maximum delay spread. The statistics of these parameters are presented. Also, the coherence bandwidth of power line channels is studied and its relationship with the rms delay spread is developed. It is in view of this work that further research in power line communication and related topics shall be inspired.