
WINDOWS STRAIGHT
-60%
DETECTION REACHES
+100%
DETECTS MOST
+5%
Imagine you and fifteen friends are standing in a field, all looking at a distant mountain. You all agree on where the mountain is. Now someone puts a flashlight on a pole 5 km away and tries to convince you it's the mountain. From YOUR angle the flashlight looks slightly left; from your friend 90 meters away it looks slightly right. That angular disagreement — parallax — is something a real mountain 20,000 km up would never produce. This paper turns that geometric insight into a spoofing detector for drone swarms. The specific threat is GNSS-WASP, a recently proposed attack that places one transmitter per spoofed satellite a few kilometres from a drone swarm. The attacker carefully times signals so every drone's computed position shifts by the same offset, preserving the swarm's formation. Existing formation-based cross-checks — the main cooperative defense — see nothing wrong because every drone's relative position looks correct. GNSS-WASP's own authors proposed a countermeasure based on carrier-phase double-differences, but this paper shows it catches the attack in at most 5% of time windows during straight flight. Park and Yoo's key observation is that parallax is inescapable when the transmitter is close. For a 90 m swarm and a transmitter at 5 km, the pseudorange residuals bend by 20-40 cm across the formation — far below position-level detection but well within the measurement precision of standard GNSS receivers. Crucially, this signature persists regardless of whatever time delays the attacker injects, because the attacker cannot change the physical geometry between their transmitters and the drones. The detector is a generalized likelihood ratio test (GLRT) applied to the pseudorange and Doppler measurements every drone already collects. All clock biases and attacker-introduced delays are treated as unknown nuisance parameters and profiled out. No extra antennas, no specialized hardware, no inter-drone ranging — just the standard GNSS observables fed into a statistical test. The test checks whether the residual pattern across all drones is consistent with satellites at orbital distance or betrays a near-field source. Simulations use real GPS and Galileo almanac orbits with white Gaussian receiver noise. Sixteen drones in a 90 m formation detect GNSS-WASP at 5 km range in 13-60% of 10-second windows during straight flight, depending on noise level and constellation. During formation turns — which create Doppler diversity — detection jumps to 100% with a precise Doppler receiver. The same test framework also catches single-transmitter meaconing and ground-based antenna-array spoofing, making it a general near-field detector rather than a GNSS-WASP-specific patch. The honest limitation is that detection probability during straight flight is modest: 13-60% per window means you need multiple windows to reach high cumulative confidence, and a clever attacker who times the spoof to straight-line segments could exploit this. The paper acknowledges this directly and shows that any maneuver — even a gentle turn — dramatically improves detection through Doppler diversity. The practical implication is that a swarm running this test should occasionally execute small formation maneuvers specifically to generate detection-friendly geometry. This is a clean, physics-grounded contribution to cooperative GNSS integrity. It does not require exotic hardware or impractical coordination. The gap between simulation and field demonstration is the obvious next step, and the paper is transparent about operating entirely in simulation with idealized noise models. But the geometric argument is sound: parallax from near-field transmitters is a physical invariant that no signal-layer trick can erase.
The analogy
You and fifteen friends are looking at a distant mountain. Someone plants a flashlight on a pole 5 km away to impersonate it. From your position it's slightly left; from your friend 90 m away it's slightly right. A real mountain 20,000 km up would never produce that angular disagreement. This paper turns that parallax — unavoidable when the source is close — into a spoofing detector for drone swarms using measurements they already collect.
Flashlight trick
A close light source creates different viewing angles for observers spread across a field — parallax reveals its proximity
GNSS parallax test
Near-field spoof transmitters create pseudorange/Doppler residuals across a drone swarm that betray their proximity vs orbital satellites
Honest assessment
A clean, physics-grounded contribution that identifies a real gap in cooperative GNSS anti-spoofing (formation-preserving attacks slip through existing checks) and proposes a principled detector using measurements drones already collect. The geometric argument is sound and the numbers beat the only existing defense decisively during maneuvers. The limitation is entirely in validation: simulation-only with idealized noise, no field test, no adaptive attacker. This is a strong simulation-stage paper that earns a field demonstration as its next step, not a finished capability.
What Is CLAIMS?
70
/ 100
BedrockNews applies a structured analytical framework called CLAIMS to academic papers and technical research releases.
How CLAIMS worksClaim
3/5
Ladder
4/5
Architecture
4/5
Integrity
2/5
Milestone
4/5
Successor
4/5
What is this paper actually committing to?
Real but incremental. Parallax as a spoofing signature is a sound physical insight applied to a specific new threat (GNSS-WASP). Not first-of-kind anti-spoofing, but first to exploit near-field parallax across a swarm cooperatively.
The field fight
Can formation-preserving GNSS spoofing be detected without extra hardware beyond standard GNSS receivers?
Side A
Carrier-phase double-differences and formation cross-checks are sufficient for cooperative spoofing detection
Held by: GNSS-WASP countermeasure, prior formation-check literature
Side B
Physical-layer geometric tests (parallax, Doppler diversity) are necessary because signal-layer tricks preserve formation consistency
Held by: This paper
Milestone tracker
Unlocks: Operational anti-spoofing capability for military and commercial swarms without hardware modifications
Current
16
drones (simulation)
this paper
Next
4-16
drones (field test)
real hardware validation
~1-3 years with funded team and regulatory access
Verified Facts (7)
Parallax bends signal arrival times across a 90 m swarm by 20-40 cm for a transmitter at 5 km
16 drones detect GNSS-WASP at 5 km in 13-60% of 10 s windows in straight flight
Detection reaches 100% during formation turns with a precise Doppler receiver
GNSS-WASP countermeasure detects at most 5%
The test catches single-transmitter and ground-array spoofing
Simulations use real GPS and Galileo orbits with white receiver noise
16 pages, 7 main figures, 7 main tables
Practical Implications (4)
Drone swarm operators can add this GLRT as a software-only spoofing check using existing GNSS observables — no hardware changes required
Swarms should incorporate periodic small formation maneuvers to maximize Doppler diversity and push detection toward 100%
Defense planners should not rely solely on formation cross-checks against sophisticated spoofing — GNSS-WASP-class attacks evade them
The 20-40 cm parallax signature at 5 km sets a minimum attacker standoff distance; pushing transmitters beyond ~20 km likely defeats this test, defining the operational envelope
More latest news
Read all