928 IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, VOL. 60, NO. 3, MARCH 2011 Wireless Sensing of Human Respiratory Parameters by Low-Power Ultrawideband Impulse Radio Radar Joshua Chong Yue Lai, Ying Xu, Erry Gunawan, Member, IEEE, Eric Chern-Pin Chua, Arash Maskooki, Yong Liang Guan, Kay-Soon Low, Senior Member, IEEE, Cheong Boon Soh, Senior Member, IEEE, and Chueh-Loo Poh, Member, IEEE Abstract—The simultaneous tracking of the chest respiratory rate and amplitude of human beings using a low-power ultra- wideband (UWB) impulse radio signal is investigated for the application of sleep apnea monitoring. The measurement of res- piratory amplitude, in addition to the breathing rate, is crucial for sleep apnea assessment which requires, among other things, the accurate estimation of tidal volume per minute. Backscattered UWB impulse radio signals from a human subject are detected in the time domain to calculate the chest displacements in this paper. Since the pulse disposition in time is linearly related to the chest movement, the amplitude of the chest movement can be extracted accurately without calculation approximations normally used in many of the existing methods. The multipeak detection of the pulse disposition method, instead of using only the single pulse peak detection, is also proposed to improve the accuracy of the calculation further. The experiments are carried out on four human subjects with different sizes and genders. The correlation of the chest movements’ amplitude and breathing rate between the simultaneously measured results obtained by our method and by the respiratory chest band is very good. Index Terms—Impulse radio radar, respiratory parameters, sleep apnea, ultrawideband (UWB), wireless sensor. I. I NTRODUCTION U LTRAWIDEBAND (UWB) radio engineering has long been investigated for the noncontact detection of respi- ration [1]–[3]. One important application of the noncontact wideband radio sensor is in the assessment of obstructive sleep apnea syndrome (OSAS), which is a breathing disorder charac- terized by the repetitive obstructions of the upper airway which lead to complete or partial airflow cessation. Polysomnography is the current standard diagnostic modality of OSAS, but it is relatively inconvenient, expensive, and inefficient [4]. The noncontact wideband radio sensor therefore presents a signif- icant advance, as it significantly redefines the notion of non- intrusiveness and patient-friendliness for home-based usage. It is not affected by clothes or blankets and is also safe to the Manuscript received January 18, 2010; revised April 30, 2010; accepted June 30, 2010. Date of publication November 9, 2010; date of current version February 9, 2011. The Associate Editor coordinating the review process for this paper was Dr. Devendra Misra. J. C. Y. Lai, E. Gunawan, A. Maskooki, Y. L. Guan, K.-S. Low, C. B. Soh, and C.-L. Poh are with the Nanyang Technological University, Singapore 639798. Y. Xu is with TDK-EPC, Singapore 349249. E. C.-P. Chua is with the Duke-NUS Graduate Medical School, Singapore 169547. Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TIM.2010.2064370 environment, including the human body, due to the very low electromagnetic energy emission. Current research on noncontact respiration movement de- tection focuses on a Doppler radar technique based on carrier phase detection or changes in the wavelength when radio waves reflect off the moving chest wall [5]–[7]. However, the key disadvantage of this technique is the inaccurate measurements of the respiratory amplitude. This is due to an inherent approxi- mation in a key processing step and to its sensitivity to ambient conditions, which necessitates constant recalibration. The development of the carrierless UWB impulse radio (UWB-IR) systems has provided new means for the noncontact measurement of the human respiration [8]. In this method, a train of subnanosecond pulses is transmitted toward a subject, a vector network analyser is used as the receiver to measure the frequency response of the received backscattered signal, and, then, an inverse fast Fourier transform is taken to derive the power delay profile of the signal in the time domain. This paper shows the feasibility of measuring the respiratory rate accurately. However, the frequency domain measurement method does not give a direct measurement of the chest move- ment amplitude, and the inherent approximation used in taking the inverse Fourier transform to derive the signal in the time domain will lower the accuracy of the respiratory amplitude measurement. The measurement of the respiratory amplitude is crucial for the OSAS assessment which requires, among other things, the accurate estimation of the tidal volume per minute. In this paper, we propose an accurate and straightforward method to detect both the respiratory amplitude and rate using the carrierless UWB-IR system which complies with the Federal Communications Commission (FCC) regulation on electro- magnetic emission [9]. Instead of measuring the frequency response of the received backscattered signal, the proposed method measures the pulse disposition in the time domain which is linearly related to the chest movement, and hence, the absolute amplitude of the chest movement can be ex- tracted accurately. The multipeak detection of the pulse dispo- sition, instead of using only the single pulse peak detection, is also proposed to improve the accuracy of the calculation further. Several different scenarios, i.e., with the subject in sitting and sleeping positions, are investigated. The measure- ment results obtained using our proposed method are then compared with the simultaneous measurement results obtained using a respiratory band which measures the chest movements directly. 0018-9456/$26.00 © 2010 IEEE