Study On Channel Estimation Technique Base On Homeplug Green PHY System

Power Line Communication(PLC) refers to the data transmission and information exchange through power line. With the development of modulation technology, transmission technology and signal processing technology, Power Line Communication technology is also developing rapidly. The transmission rate has increased from Kbps into Gbps. Thanks to its low cost, no need of rewiring, wide coverage and easy access, PLC has become one of the competitive solutions for broadband access.Starting from the basic background of PLC, this thesis made a brief introduction of the development and the state of the art of the PLC in China and abroad. The interpretation of the physical layer structure in accordance with the Homeplug Green PHY protocol laid a foundation for further research and simulation platform. Then the research on broadband PLC channel and channel noise were introduced, the channel and noise models with experimental measures and data collection are represented to elaborate the characteristics of the PLC system.The thesis further explained the OFDM system and channel estimation methods. After a meticulous research based on existing methods and optimization, a method based on the MMSE criteria in the frequency domain by exploiting the MDL and ESPRIT algorithms is proposed. To verify the proposed method, this algorithm is then realized in a Homeplug Green PHY system which is built on C and Matlab platforms. The simulation results show that the proposed MDL and ESPRIT based algorithm can improve the overall performance of PLC system, especially under an unknown channel.