Networked staring radar testbed for urban surveillance: Status and preliminary results

Mapping and characterising low to medium airspace in an urban setting using radar presents significant challenges, especially for low observable targets such as drones and birds. The task of detecting, locating, tracking and identifying such small, low and slow targets, in dense clutter requires adaptation of both radar hardware and the processing algorithms. We are creating a testbed of two L-band staring radars at University of Birmingham (UoB) to understand how novel radar architectures help us to meet these challenges. Each radar uses a floodlight beam on transmit and has a fully digitised 64-element receiver array. The fully flexible beam-forming and persistent dwell enable high sensitivity and fine Doppler that is sufficient to detect and classify small drones out to a range of several kilometres. The testbed will enable the collection of extensive datasets to support machine learning techniques for classification of targets along with long-term monitoring and mapping of bird activities in an urban environment. An ultra-low phase noise Quantum oscillator will be used to replace the conventional oscillator in the radar and also act as a master clock to synchronise the two systems. The testbed will be used to investigate monostatic, bistatic and multistatic processing techniques and thereby provide an invaluable resource for supporting research in a range of industrial applications for ultra-low phase noise multistatic radar, and to showcase these in operationally realistic conditions. This paper highlights some of the challenges in installing the radar at the UoB campus and initial benchmarking results from field trials are presented from both radar systems that are now fully operational.