Microwave field detection and recognition technology, leveraging its safety features such as noncontact and nonradiation characteristics, has achieved widespread application in medical diagnosis, industrial nondestructive testing, and public security screening. However, high latency from antennas, multiple analog-to-digital conversions, and the serial operation mode of sensing/storage/computation result in significant system delays, preventing the capture and recognition of highly dynamic microwave information and leading to low operational efficiency. Here, we construct an NV color-center in-sensor computing microwave sensor (NV-ISCMS) using a diamond array. Leveraging the high linear correlation between the electron spin resonance intensity of NV color centers and microwave power, ultrafast detection of microwave fields is achieved. Controllable adjustment of the device operating frequency is realized based on the Zeeman splitting resonance frequency shift effect. Each diamond, regulated by a positive–negative voltage follower circuit, serves as a tunable responsivity NV color-center microwave sensing unit. By implementing an in-sensor computing architecture that performs matrix multiplication of microwave field intensity/responsivity and Kirchhoff’s law current summation, parallel microwave sensing and data processing are achieved without introducing any analog-to-digital conversion processes. Experimental verification shows that the device’s single microwave field target detection and recognition time is only 153.2 ns, providing a reliable technical solution for achieving fast response, low power consumption, and reduced hardware overhead in intelligent microwave field detection and recognition technology.
Gao et al. (Mon,) studied this question.