呼吸作為生理系統的一項特徵，被視為重要指標用於評斷睡眠及心血管系統之健康與否。傳 統用於呼吸偵測的技術主要是透過睡眠檢查系統(Polysomnography, PSG)來進行，PSG系統造價 高，通常配置於醫院的睡眠中心，屬於接觸式裝置，使用者易感受到不適感。本研究說明如何利 用都卜勒雷達進行遠端呼吸訊號量測， 系統使用的天線為EUSSO-UWA2614-OP， 訊號頻段為 2.4GHz，與一般的家用無線路由器相同。而此方法主要的困難點在於如何在一公尺的距離範圍內 精確地量測呼吸，我們設計一套演算法分析及過濾接收到的訊號，並從中找出呼吸模式藉此推估 呼吸率，精確率可達90%。

Respiration is an important feature of human physiological system for estimating the quality of sleep and the health of cardiovascular system. Polysomnography (PSG) traditionally used in sleep center of hospital for the research of sleep, but the cost is prohibitive. Patients easily feel uncomfortable in diagnostic process because they have to equip with different sensors on body (ECG, EKG…etc). We design a doppler-radar-based system which can detect respiration remotely. The system uses antenna EUSSO-UWA2614-OP to receive breathing signal. Its baseband is 2.4GHz, which is similar to the one in a home use wireless AP. The main challenge we would like to solve is how to detect respiration precisely in a range of 1 meter. We design an algorithm to analyze and filter the received breathing signal, and then find the respiration pattern and estimate the respiration rate accordingly.