1558-1748 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2018.2873824, IEEE Sensors Journal 1 Abstract—Wingbeat frequency (WBF) is an important biological feature for flying birds, used for species classification and activity observation. To develop a handheld system for WBF detection, a 2.4-GHz, short-range continuous-wave (CW) radar system is proposed in this work. By utilizing the self-injection locked (SIL) radar architecture and joint time-frequency analysis (JTFA), this system is able to detect WBF on a flying bird from a distance up to 18 m. Several experiments, including identifying species in a mixed group and distinguishing the target from the clutter, are demonstrated with bionic birds. The measurements with real flying birds are also conducted in both indoor and outdoor scenarios. This proposed radar system can be used in the laboratory environment or in the wild, showing a great potential for biological study. Index Terms—wingbeat frequency (WBF), biological, bird, radar, self-injection locked (SIL) I. INTRODUCTION HERE are nearly a million species of flying insects and of 13000 flying warm-blooded vertebrates, including 9000 birds, 1000 bats and other mammals. They represent a dominant component of biodiversity. With such a significant number of species and quantity, in many cases the activities of flying species can have significant impacts on the ecosystems. Especially birds, which are located at the top of ecological pyramid, function as pollinators, predators, scavengers, and seed dispersers. They form an important part of the food chain. By contributing tremendously to the health of the ecosystem, these flying birds also offer a number of direct benefits to human society, such as agriculture and the economy. Furthermore, since those flying species are very sensitive to the Manuscript received July 19, revised Sept. 4, 2018. This work was supported in part by the Ministry of Science and Technology, Taiwan under Grant 106-2221-E-194 -029. C.-C. Chang is with Department of Electrical Engineering and the Department of Communications Engineering, Center for Telecommunication Research, National Chung-Cheng University, Chiayi 62102, Taiwan (phone: +886-5-272-0411#33217; e-mail: ccchang@ee.ccu.edu.tw ). Y.-S. Su is with the Department of Electrical Engineering, National Chung-Cheng University, Chiayi 62102, Taiwan (e-mail: yusheng0509@gmail.com ). J.-C. Kuo, was with Department of Electrical Engineering, National Chung-Cheng University, He is now with the Novatek Microelectronics Corporation, Hsinchu 30076, Taiwan (e-mail: jim771212@gmail.com ). J.-M. Kuo, was with Department of Electrical Engineering, National Chung-Cheng University, He is now with Gemtek Technology Corporation, Hsinchu 30352, Taiwan (e-mail: kjmings@gmail.com ). . TABLE I EXAMPLES OF FLYING BIRDS AND THEIR WBFS WITH CORRELATED WING PARAMETERS. Bird Mass (kg) Wing span (m) Wing area (m 2 ) Wing beat frequency (Hz) Brambling 0.0228 0.262 0.013 18.2±1.74 Buteo 0.964 1.29 0.254 3.63±0.168 Starling 0.0884 0.384 0.0251 10.6±1.25 Wigeon 0.77 0.822 0.0829 6.83±0.33 Wood Pigeon 0.495 0.751 0.0797 ~6.8 Cape Pigeon 0.418 0.875 0.0773 5.56±0.256 Seagull 0.364 1.1 0.138 2.98±0.151 * values taken from the literature [5]-[6], and may vary with different breeds and species living environment, they can be used as the prior indicators on weather and landscape changes. Accordingly, the study of birds has fascinated physicists and biologists for decades. Scientists’ classification can help to clearly identify species, then study and observe them. Birds can be classified based on several biological features, including body size and shape, coloration, voice, behavior patterns, and wingbeat frequency (WBF). Various techniques have been introduced to extract different features, but most of them depend on visual and auditory observations [1]-[4]. Over the past decades, the ecologists have started to investigate the importance of WBF. WBF is measured as the number of flapping cycles per second, which can be very diverse for different species due to the variety in body size, wing structure and muscular components. Since it highly depends on physical parameters, for those species having similar mass but different wing morphology may result in different WBFs. Even the same species but different breeds, the WBF can be different due to the body and wing size variation. Table I gives several examples on birds with their corresponded WBFs [5]-[6]. Note that WBFs are generally higher in small flies than in big flies. Except classification, WBF can also reflect the behavior change of fly species. For example, it has been reported that migratory birds have the highest WBF during initial ascent, but the lowest WBF during final descent [7]. Honeybee is another example that WBF is reduced after pollen collection due to weight gain. Therefore, WBF can be used to observe those species’ behaviors as well as to monitor their health statuses. Alternatively, WBF is also related to the flying conditions, including temperature, gravity, wind speed, humidity and air density [5]. As a consequence, it can be used to estimate the weather condition, or even to predict the migration spread since it is correlated to the expenditure of energy during flight. Remotely sensing bird’s WBF was first reported using radio Design of Self-Injection-Locked Radar System for Birds’ Wingbeat Frequency Detection Yu-Sheng Su, Chia-Chan Chang, Member, IEEE, Jer-Chun Kuo and Jia-Ming Kuo T