電力線通訊(PLC)系統是以室內電力線網路為主要的傳輸媒介，因其具有覆蓋範圍廣、不用重新佈線、連接方便等優點，所以在近年來被世界各國所重視。雖然電力線通訊具備最後一哩的優勢，但是因為傳輸環境的雜訊干擾以及通道衰減嚴重，使得在實際資料傳輸率方面與理想的情況會有很大的落差。因此如何解決電力線傳輸通道衰減以及抵抗雜訊干擾以提升傳輸效率，一直是各界研究的重點所在。 本論文提出了指數型增益的自動增益控制與載波回覆(AGC-CR)和可調整增益的數位座標軸旋轉計算器(MGC-CORDIC)來對衰減通道做等化與補償。自動增益控制與載波回覆為將通道衰減分成增益和相角兩部份，利用LMS 演算法做適應性補償，但是這種處理方式在增益的部分須另外處理。而所提出指數型增益係數的自動增益控制與載波回覆則不須再行處理，而且還可以增快收斂的速度。 可調整增益的數位座標軸旋轉計算器則是根據輸入向量的大小和相角來計算出通道衰減程度以及相位的旋轉，可以用來取代LS 通道估測方法中的除法器。最後我們使用具有近似指數型增益的係數，配合指數型增益係數的自動增益控制與載波回覆來降低硬體複雜度。 Power line communication (PLC) system is an indoor wire-line communication system which adopts the power line network as the transmission medium. Since the coverage area of power line network is widespread in the world, the infrastructure of power line does not need be rebuild again. Accordingly, it has the convenient advantages to be the last mile system and is paid a great attention by many countries in recent years. Although the power line communication has superiority in the last mile, the noise influence and channel fading on transmission environment are still very serious. There is still an obvious drop between the real and ideal transmission rate. Therefore, how to mitigate channel fading and resist noise influence is the key point to increase transmission efficiency. This thesis proposes an Auto-Gain Control and Carrier Recovery (AGC-CR) with exponential gain and a Modified Gain Control - COordinate Rotation DIgital Computer (MGC-CORDIC) to compensate for channel fading. Traditionally, AGC-CR divides channel fading into gain and phase angle, and use LMS method to make the phase adaptive compensation. But it needs a more complex gain processing additionally. On the other hand, the proposed AGC-CR doesn’t need. Furthermore, the proposed AGC-CR can speed up the converging rate. MGC-CORDIC calculates channel fading and phase rotation according to the magnitude and phase angle of the input vector. It can replace the function of divider in the Least Square (LS) estimation method. Finally we combine MGC-CORDIC and the proposed AGC-CR to reduce the hardware complexity.
