Sleep is important to maintain human's physiological and psychological well-being. The quality is often affected by many environmental factors, such as air dust, temperature and noise. This paper proposes a tent-type clean unit system platform (T-CUSP) to construct a high air quality sleep environment. In this approximate independent-system, the feasibility of non-contact sleep monitoring method based on gas-molecule and air-particle analysis would be evaluated. In this paper, 7 overnight sleep experiments were performed with simultaneous recording of mini PSG (EEG, EOG, and EMG) for sleep staging and actigraph for body movement. The air quality was monitored by the particle counter and fluctuation of air-borne particle. The average concentrations of PM2.5, PM10, and CO2 in the T-CUSP during sleep were 4.85±3.2 μ g/m3, 8.43±4.5 μ g/m3, and 1122±128 ppm, respectively. The air-quality of T-CUSP conform to the agreed limits in US and EU. Furthermore, the fluctuation of CO2 can reflect the behavior of subjects. 91% of CO2 rising epochs were accompanied by a significant body movement and could be identified as the "wake" stage. The short experimental results demonstrate the feasibility of constructing an air-clean environment with non-contact sleep assessment to support healthy sleep.