近年來人們逐漸依賴消費性電子產品,這些產品也漸漸融入家庭生活中。家庭數位化是一種趨勢,然而產品間的通訊需要大量的重新佈線,對於家庭來說是一種麻煩,這使得電力線擁有了不必重新佈線的優勢,且每個家庭都有完善的電源設施。 本文提出利用小波轉換先對解耦合後之訊號,進行濾除雜訊的動作,再進入X10解調變判斷訊號的數位資料,最後再進行錯誤率的統計。本文利用Verilog硬體描述語言設計為晶片,以FPGA來驗證功能。我們針對電力線通道的問題,設計三個架構來統計調變的傳輸錯誤率,分別是改變電力線長度、改變電力線負載和改變電力線雜訊三種電力線通道來統計傳輸錯誤率。 本文以上述電力線通道進行實作,其結果發現X10調變在改變電力線長度與改變電力線負載上,傳輸錯誤率並無明顯的差異,對電力線雜訊有著明顯的影響。而加入小波轉換多重解析度分析來濾除雜訊後,在實作結果中也顯示,雜訊對於X10調變的影響已經大大降低。
Nowadays, people use many consumer products which become parts of house life. Smart home is a hot trend; however, communication lines need to be reconstructed. The power line becomes the last mile for achieving the smart home technology because it is not necessary to re-layout the electricity wire. This thesis used the wavelet transform to filter the noise in the power line communication before the signal is modulated by X10. Then the accuracy rate is evaluated. This thesis used Verilog hardware language to design the proposed method on an FPGA chip. There are three topologies developed in this thesis to investigate the power line channel: various lengths of the power line, various loading of the power line, and various SNRs. It is found that the accuracy rates of transmittal signals are independent of lengths and loadings of the power lines, compared with SNRs. Wavelet transform plays an important role to reduce the noise in the power lines and increase the accuracy rate.