Background
Passive imaging detects naturally emitted or reflected electromagnetic radiation from objects without relying on active signal transmission. This approach spans multiple spectral bands—including visible light, infrared (IR), and millimeter-wave (mm-wave)—and offers key benefits such as covert operation, low power consumption, and simplified system architectures. These attributes make passive imaging particularly attractive for low-cost, compact systems where thermal or material-specific information is desirable.
IR-based passive imaging is widely deployed; however, it is primarily effective at night and is limited in its ability to penetrate materials like walls and glass. In contrast, mm-wave and terahertz (THz) passive imaging technologies have gained increasing attention due to their superior material penetration capabilities and robustness under diverse environmental conditions. These properties enable applications such as concealed weapon detection, security screening, and non-invasive medical diagnostics.
A key advantage of mm-wave and THz imaging lies in the existence of atmospheric transmission windows at frequencies such as 35, 94, 140, and 220 GHz, where signal attenuation remains low—even in challenging environments involving fog, dust, or precipitation. While III-V compound semiconductor technologies have been traditionally used in these systems due to their high-performance characteristics, there is a growing shift toward CMOS implementations. Despite the higher flicker noise compared to SiGe technologies, CMOS is increasingly favored for its low power consumption, scalability, and excellent integration capabilities, enabling the development of compact and cost-effective passive imaging solutions.
Our Approach
The Dicke-SW-based passive imager sets a new benchmark by achieving record-low power consumption without compromising state-of-the-art performance. A major breakthrough in this prototype is the first-ever demonstration of noise suppression in a passive imager, addressing a critical challenge in the field. Additionally, we have developed a comprehensive theoretical noise suppression analysis and holistic noise suppression scheme, incorporating system-level, circuit-level, and device-level non-ideality effects, to provide deeper insights into noise behavior and enable further optimization of passive imaging systems.
We are developing a new mm-Wave/THz passive imaging architecture that offers several key advantages over conventional designs:
- High immunity to clock duty cycle distortion, improving system reliability and performance consistency.
- 100% sensing time, maximizing data acquisition efficiency and enhancing detection capabilities.
- Well-matched signal and reference paths, reducing errors and improving image fidelity.
By integrating these advancements, this work will establish a new paradigm for low-power, high-performance passive imaging, enabling more robust and efficient applications in security, surveillance, biomedical imaging, and remote sensing.
References
- Q. J. Gu et al., “A CMOS Integrated W-band Passive Imager,” in IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 59, no. 11, pp. 736-740, Nov. 2012.
- Q. J. Gu, Z. Xu, H. -Y. Jian and M. -C. F. Chang, “A CMOS fully differential W-band passive imager with <2 K NETD,” 2011 IEEE Radio Frequency Integrated Circuits Symposium, Baltimore, MD, USA, 2011, pp. 1-4, doi: 10.1109/RFIC.2011.5940669.
- Q. J. Gu et al., “A 100 GHz Integrated CMOS Passive Imager with >100MV/W Responsivity, NEP,” IET Electronics Letter, vol. 47, issue 9, 2011, Featured Paper, pp. 544–45.
- A. J. Tang, Y. Kim and Q. J. Gu, “25.4 A 0.43K-noise-equivalent-ΔT 100GHz dicke-free radiometer with 100% time efficiency in 65nm CMOS,” 2016 IEEE International Solid-State Circuits Conference (ISSCC), San Francisco, CA, USA, 2016, pp. 430-431, doi: 10.1109/ISSCC.2016.7418091.
- Z. Mohseni, S. Sabbaghi, H. Yu, P. Han, and Q. J. Gu “D-band Dicke switch based Passive Imager with 0.13K NETD in 28nm CMOS Technology,” IEEE CICC2025