雷達的型式分為連續波雷達、脈波雷達、與壓縮雷達。連續波雷達依調變的不同又分為三種，即振幅調變、相位調變、與頻率調變，其中頻率調變分為線性與弦波兩種。脈波雷達分為低脈波重複週期、高脈波重複週期、與多重脈波重複週期三種。而壓縮雷達分為頻擾雷達與編碼雷達兩種。 本篇論文研究動機主要是討論多重脈波重複週期訊號與擾頻雷達之頻率訊號經過一個特有裝置後，訊號的型式已與原訊號有所不同，經過此裝置後的訊號仍須由人工來作訊號辨識的工作，而本論文的目的則是寫出一套自動辨識軟體來改進人工辨識方法。首先將訊號做一些類型的分類，利用最大相似性(ML)估測法則作為訊號判斷的方法，之後採用電腦模擬作為訊號辨識的效能評估工具。 Radar could be divided into CW radar, pulse radar, and compressed radar. CW radar could be further divided into AM-type, PM-type, and FM-type based on the choice of modulation. FM could be linear or sinusoidal. Pulse radars could be divided into low-pulse repetition interval radar, high-pulse repetition interval radar, and multiple-pulse repetition interval radar. Compressed radar could be divided into chirp radar and coded radar. When the signals of the multiple-pulse repetition interval radar and the frequency of chirp radar go through a specific instrument, the output signals are different from the original ones. The output signals still need to be recognized by man. This thesis submits an automatic solution to solve this problem. The first step is to classify the signal, and then use Maximal-likelihood criterion to estimate the signal. We use the computer simulation to evaluate the efficiency of signal recognition.
