公共交換電信網(PSTN)是目前主流的個人通信系統。此外,專用交換機(PBX)在企業的通信系統中也是很受歡迎的小型交換設備。所有的公共交換電信網、專用交換機與電話都是模擬和使用DTMF信號,與對方互相溝通傳遞訊息。因此,在公共交換電信網、專用交換機與其他電信設備中,DTMF訊號檢測是一個重要課題。 本文提出一種新式DTMF偵測方法,此方法是改進離散傳利葉轉換(DFT)演算法,提出使用最佳相位有效降低DTMF偵測之計算量。其最佳相位是使用Open-loop搜尋方法獲得。比過去傳統的離散傅利葉轉換(DFT)或Goertzel演算法,不只能有效地降低75%至83%的計算量,也能有效地保持頻譜的效率,由實驗結果中可看出新式的偵測方法是更加優異於傳統的演算法。
Public switch telecommunication networks (PSTNs) are often used in personal communication systems (PCSs). A private branch exchange (PBX) is also popular for business communication systems. PSTNs, PBXs, and phone sets are analog and use a dual-tone multi-frequency (DTMF) signal to communicate with each other. Thus, DTMF signal detection is important in PSTN and PBX systems. This dissertation presents a new algorithm for detecting DTMF signals. Discrete multi-frequency transform (DMFT) algorithm with an optimal phase is used to reduce the DTMF detection calculations. The DMFT approach uses a pass-band ripple as a cost function, and the optimal phase is obtained using an open-loop search method. Compared to the traditional discrete Fourier transform approach, the proposed method achieves computational savings of 75% to 83% and maintains spectrum efficiency. The experimental results show that the DMFT algorithm performs exceptionally.