A Passive Software-Defined Radio-based mmWave Sensing System for Blind Integrated Communication and Sensing

TL;DR

This work tackles regulatory and synchronization barriers in ISAC by proposing a fully passive mmWave sensing system that operates solely on ambient signals. It deploys a differential two-channel SDR receiver with a common local oscillator to cancel the unknown transmitted waveform and common distortions, formulating the relative channel $H_{ ext{rel}}(f,t)=\frac{H_2(f,t)}{H_1(f,t)}$ to isolate motion-induced dynamics. A complete signal-processing chain—frame segmentation, differential channel computation with spectral averaging, and NU-STFT Doppler estimation—extracts robust Doppler signatures from non-cooperative ambient signals. Experimental validation at 25.1 GHz with a commercial transmitter and COTS SDR confirms accurate Doppler measurements for a controlled plate and diverse human activities, including multi-person scenarios, demonstrating the method's potential for blind ISAC in real-world deployments.

Abstract

Integrated Sensing and Communication (ISAC) is considered as a key component of future 6G technologies, especially in the millimeter-wave (mmWave) bands. Recently, the performances of ISAC were experimentally evaluated and demonstrated in various scenarios by developing ISAC systems. These systems generally consist of coherent transmitting (Tx) and receiving (Rx) modules. However, actively transmitting radio waves for experiments is not easy due to regulatory restrictions of radio. Meanwhile, the Tx/Rx should be synchronized and Rx need the information of Tx. In this paper, a fully passive mmWave sensing system is developed with software-defined radio for blind ISAC. It only consists of a passive Rx module which does not depend on the Tx. Since the proposed system is not synchronized with Tx and has no knowledge of the transmitted signals, a differential structure with two oppositely-oriented receivers is introduced to realize the sensing function. This structure can mitigate the influences of unknown source signals and other distortions. With the proposed sensing system, the ambient mmWave communication signals are leveraged for sensing without interrupting the existing systems. It can be deployed for field applications such as signal detection and dynamic human activity recognition since it does not emit signals. The efficacy of the developed system is first verified with a metallic plate with known motion pattern. The measured Doppler spectrogram shows good agreement with the simulation results, demonstrating the correctness of the sensing results. Further, the system is evaluated in complex scenarios, including handwaving, single- and multi-person motion detection. The sensing results successfully reflect the corresponding motions, demonstrating that the proposed sensing system can be utilized for blind ISAC in various applications.

Figures (12)

• Figure 1: Schematic diagram of the proposed passive mmWave ISAC sensing system.
• Figure 2: Deployment geometry of the differential architecture for passive ISAC sensing. (Black lines: line-of-sight (LOS) and other static MPC to Rx1. Red lines: dynamic component from moving target to Rx2. Blue lines: other static MPCs such as reflection from wall to Rx2.)
• Figure 3: Illustration of aligned frames used for analysis. Pink part is the uniform window size.
• Figure 4: Relative channel response $H_\text{rel}(f_q, t_i)$ in time-frequency domain. (a) Concept diagram. (b) Measurement result.
• Figure 5: Experimental hardware and validation setup. (a) A close-up view of the proposed system in operation. (b) The controlled validation setup, featuring a metallic plate mounted on a linear slider equipped with a stepping motor.