Analysis Of Channel And Noise Over Power Line Based On Compressive Sensing

The power line communication (PLC) has been a topic of interest in recent years. Significant research efforts and works in industry and academia have been carried out to model the communication channel and the noise over PLC channel, providing theoretical basis for PLC equipment research and development. However, few studies concentrate on the characteristic of the source data of the channel and the noise. When the PLC equipments communicate with each other with a bandwidth up to a few MHz, the source data will consume large amounts of storage and computing resources. Therefore, a reasonable data compression and storage solution is very important and necessary.This paper researches on designing a efficient data storage and collection strategy of PLC channel and noise source data based on compressed sensing.Firstly, the noise over the PLC channel can be divided into two categories, namely background noise and random noise, under a certain threshold based on the analysis of the sparse characteristics. On the one hand, the random noise can be compressed to store based on a novel data compression technology, compressed sensing, because of the sparse characteristic in the frequency domain of it. On the other hand, background noise is similar to white noise, so we can recover it through two parameters, stored noise mean and variance. Extensive simulations are given to analyses the performance of the proposed scheme.Secondly, similar to wireless multipath channel, the PLC channel model suggests that the characteristics of it are mainly depend on short distance and small delay paths. As a result, the compressed sensing can also be applied to PLC channel estimation. With a appropriate pilot signal on the transmitter, the receiver receive and store main features of the channel parameters, which can be used to retrieval the power line channel characteristics anytime and anywhere. Moreover, the simulation is given to verify the proposed scheme.