SkyWatch: A Passive Multistatic Radar Network for the Measurement of Object Position and Velocity

Brief

Randall et al. (2023) describe SkyWatch, a passive radar network that repurposes commercial FM broadcast stations (88–108 MHz) as transmitters of opportunity, with software-defined receivers feeding bistatic range and Doppler measurements to a central server running an Extended Kalman Filter for real-time 3D position and velocity estimation at up to 15 samples per second. The system targets horizontal ranges to 150 km and altitudes to 80 km, roughly 10× the detection range of co-deployed wide-field optical and infrared cameras, which the authors calculate corresponds to approximately three orders of magnitude greater detection volume for rare events. Phase 1 testing placed six receivers around the Harvard-Smithsonian Center for Astrophysics; a prototype in Boulder, CO simultaneously tracked dozens of aircraft echoes to 120 km bistatic range. The dominant engineering constraint is direct-signal contamination: the FM transmitter's direct path arrives at ~−25 dBm, producing noise-like interference roughly 95 dB above the thermal noise floor that must be suppressed through adaptive null steering and noise canceling before aircraft-level echoes can be recovered.

Metadata

Category

Hub & Overview

Venue

Journal of Astronomical Instrumentation

Type

Peer-reviewed

Year

2023

Authors

Mitch Randall, Alex Delacroix, Carson Ezell, Ezra Kelderman, Sarah Little, Abraham Loeb, Eric Masson, Wesley Andres Watters, Richard Cloete, Abigail White

DOI

10.1142/S2251171723400044

arXiv

2305.18562

Access

Open access

Length

14 pages

Programs

Galileo Project

Instruments

SkyWatch passive multistatic radar, FM broadcast transmitters of opportunity (88–108 MHz), Beacon 8 pan-tilt security camera (8 MP), GPS-disciplined oscillators

Tags

UAP-detection, passive-radar, instrumentation, kinematics, multistatic-radar

Key points

• The system operates across the 88–108 MHz commercial FM band (wavelength ~2.8–3.4 m), sampling at up to 1 Msps, and produces 3D position and velocity estimates at up to 15 samples per second, establishing the measurement cadence for kinematic anomaly detection.p.1
• Velocity is measured directly by Doppler shift up to 2 km/s (5.8 Mach) rather than inferred from position differences, enabling first-order finite difference for acceleration, materially more accurate than the second-order differencing required by position-only systems.p.3
• At Rt = Rr = 55 km, expected received echo power from a single-engine aircraft (RCS = 15 dBsm, transmitter EIRP = 100 kW) is −116 dBm against a thermal noise floor of ~−120 dBm; 41 dB processing gain from averaging at 15 samples/s yields a nominal SNR of ~45 dB in the favorable (direct-signal-blocked) case.p.5
• The FM transmitter's direct path arrives at ~−25 dBm, generating a noise-like interference signal ~95 dB above the thermal noise floor when correlated at any non-matching delay; achieving 10 dB SNR at 15 samples/s requires ~48 dB of direct-signal attenuation.p.6
• The 150 km detection range is approximately one order of magnitude greater than co-deployed wide-field optical and IR cameras; the paper calculates this translates to roughly three orders of magnitude increased probability of intercepting a rare UAP event.p.4
• Three operational modes, direct reference, remote reference (via internet with GPS-disciplined oscillators and 1 pps timestamps for synchronization), and hybrid, allow deployment across varied terrain and transmitter geometries.p.8
• Phase 1 places six receivers around CfA: four within 30 km and two at 103 km and 148 km; three FM transmitters of opportunity are used, including one at 50 kW EIRP to the south.p.8
• Passive FM radar for UAP detection was first proposed by Davenport (1999) and refined in 2004; SkyWatch represents the first peer-reviewed multistatic implementation developed explicitly for systematic scientific UAP study.p.2

Verbatim

• “When this signal is correlated against the reference at any other delay, its phase will be randomized into a noise-like signal about 95 dB greater than the thermal noise °oor.”

  p.6

Most interesting

• The concept of passive FM radar for UAP detection was first articulated by Davenport in 1999, 24 years before this peer-reviewed implementation appeared, making SkyWatch a multi-decade idea finally reaching systematic scientific deployment.
• The FM transmitter's direct-path signal (~−25 dBm) is roughly 90 dB stronger than a typical aircraft echo (~−116 dBm), meaning the system must suppress a signal nearly ten billion times more powerful than what it is trying to detect.
• Because detection volume scales as range cubed, the 10× range advantage over optical cameras produces ~1,000× more detection volume, the paper's stated three orders of magnitude improvement in the probability of observing a rare UAP.
• SkyWatch requires no dedicated transmitter: it parasitically exploits the existing national FM broadcast infrastructure, using any station in the 88–108 MHz band as a free illuminator.
• Phase 1's 'tracking telescope' is a commercial Beacon 8 security camera (8 MP) on a pan-tilt mount, a commodity hardware choice enabling rapid, low-cost observatory deployment while a mirror-tracking telescope is readied for Phase 2.
• In remote reference mode, GPS-disciplined oscillators and 1 pps GPS timestamps are required to maintain the sub-microsecond timing precision necessary for accurate bistatic range correlation across geographically separated nodes.