FG-LiquID

Abstract

Contact-less liquid identification via wireless sensing has diverse potential applications in our daily life, such as identifying alcohol content in liquids, distinguishing spoiled and fresh milk, and even detecting water contamination. Recent works have verified the feasibility of utilizing mmWave radar to perform coarse-grained material identification, e.g., discriminating liquid and carpet. However, they do not fully exploit the sensing limits of mmWave in terms of fine-grained material classification. In this paper, we propose FG-LiquID, an accurate and robust system for fine-grained liquid identification. To achieve the desired fine granularity, FG-LiquID first focuses on the small but informative region of the mmWave spectrum, so as to extract the most discriminative features of liquids. Then we design a novel neural network, which uncovers and leverages the hidden signal patterns across multiple antennas on mmWave sensors. In this way, FG-LiquID learns to calibrate signals and finally eliminate the adverse effect of location interference caused by minor displacement/rotation of the liquid container, which ensures robust identification towards daily usage scenarios. Extensive experimental results using a custom-build prototype demonstrate that FG-LiquID can accurately distinguish 30 different liquids with an average accuracy of 97%, under 5 different scenarios. More importantly, it can discriminate quite similar liquids, such as liquors with the difference of only 1% alcohol concentration by volume.

Improving the liquid sensing resolution via multiple transceiver pairs

“…For each configuration, we repeat the experiment for 100 times. The received signal will be fed into the algorithm which proposed in [6]. Result analysis: As shown in Figure 3, when the difference in alcohol concentration exceeds 1%, the current method can achieve an accuracy of over 92%.…”

“…However, when the alcohol concentration difference is reduced to less than 0.5%, the identification rate decreases significantly. We offer a possible explanation for this phenomenon: for each receiving antenna, the amplitude of the received signal from the liquid, denoted as 𝐴 𝑙𝑖𝑞𝑢𝑖𝑑 , can be expressed using the Friis Formula [6]:…”

“…𝑑 is the distance from the radar. 𝜆 is the wavelength of the millimeter wave, and 𝑅 𝑙𝑖𝑞𝑢𝑖𝑑 presents the reflection coefficient determined by the liquid and is related to the permittivity [6].…”

“…Due to the unique permittivity of certain liquids, they can be identified by analyzing the characteristics of the reflected signal [6][7][10] [11], which leads to different attenuation of millimeter waves in different liquids. The reflected signal has a strong correlation with the material of the liquid [6], thus we can identify the liquid by analyzing the characteristics of the reflected signal.…”

“…However, due to bandwidth limitations, the sensing ability of RFID and WiFi is insufficient. In comparison, millimeter wave signals have a wider bandwidth and carry more information about the liquid, enabling fine-grained liquid identification [6] [7][8] [9]. However, current millimeter wave-based methods exhibit limited performance when it comes to identifying fine-grained liquids.…”