Passive radar for defense: bistatic detection without active emissions

A conventional radar announces itself. It emits megawatts of pulsed power, and any adversary equipped with a radar warning receiver can detect the transmission, triangulate the emitter's position, and target it with a precision strike or anti-radiation missile. Passive radar inverts this dynamic entirely: it receives but never transmits, exploiting the ocean of commercial broadcast energy already flooding the atmosphere – FM radio, digital audio broadcast, digital television, and cellular base stations – as opportunistic illuminators. The sensor is electromagnetically silent. There is nothing for an intercept receiver to detect.

This approach, formally termed passive coherent location (PCL), has moved from academic curiosity to operational deployment over the past two decades. The combination of high-performance software-defined radio receivers, GPU-accelerated signal processing, and increasingly dense commercial broadcast networks has made PCL viable as a persistent, low-cost complement to active radar systems in base air defense, border surveillance, and wide-area airspace monitoring.

The passive coherent location principle

PCL exploits the bistatic radar geometry: the transmitter (the commercial broadcast station, which PCL operators do not control) and the receiver (the PCL system) are physically separated, unlike a conventional monostatic radar where transmitter and receiver share the same aperture. When an aircraft or vehicle intercepts energy from the broadcast transmitter, a small fraction of that energy is scattered toward the PCL receiver. By comparing the received direct-path signal from the transmitter with the target echo, the system can extract bistatic range (the sum of transmitter-to-target plus target-to-receiver path lengths, minus the direct transmitter-to-receiver path), Doppler velocity, and – with multiple receiver nodes or multiple illuminators – approximate position in two or three dimensions.

The direct-path signal serves a dual purpose: it acts as the reference channel, providing the waveform replica required for matched-filter processing, and it represents the dominant interference source that must be cancelled before target echoes can be detected. Managing this reference-channel duality is the central signal processing challenge in PCL system design.

Illuminator selection: FM, DAB, DVB-T, and 5G

The choice of illuminator determines a PCL system's detection range, range resolution, and Doppler resolution – and thus which target classes it can reliably detect.

FM broadcast (87.5–108 MHz). FM transmitters are the most widely exploited PCL illuminator. Transmit powers of 10–100 kW, near-omnidirectional antenna patterns, and geographic densities of one transmitter per 30–50 km in most of Europe and North America make FM coverage essentially ubiquitous. Detection ranges of 200–300 km against large aircraft are routinely demonstrated in field trials. The significant limitation is range resolution: FM signals occupy approximately 100 kHz of bandwidth, which translates to bistatic range resolution of roughly 1,500 m – too coarse to separate small targets from ground clutter or to discriminate closely spaced objects. FM PCL is most suited to wide-area surveillance of conventional aircraft and helicopters.

DAB digital audio broadcast (174–240 MHz). DAB signals occupy approximately 1.5 MHz of instantaneous bandwidth, improving range resolution to around 100 m. The OFDM waveform structure of DAB, with well-defined pilot subcarriers, simplifies reference channel extraction and improves cross-ambiguity function quality. DAB PCL systems have demonstrated reliable detection of medium-sized aircraft at ranges up to 150 km and initial drone detections in low-clutter environments. DAB transmitter density is lower than FM in many regions, creating coverage gaps that require illuminator site surveys before system deployment.

DVB-T digital television (470–790 MHz). DVB-T provides the best range resolution of the established PCL illuminators, with 7.6 MHz channel bandwidth yielding bistatic range resolution of approximately 20 m. At UHF frequencies, propagation is more line-of-sight than FM, which reduces maximum detection range against low-altitude targets but improves the signal-to-clutter ratio in many geometries. DVB-T PCL is the current preferred illuminator for small UAV detection, where the combination of adequate range resolution and Doppler resolution allows separation of slow micro-UAVs from stationary ground clutter. Maximum detection range against large aircraft is typically 100–150 km.

5G NR (sub-6 GHz). 5G new radio base stations represent an emerging PCL illuminator with wideband characteristics (up to 100 MHz channel bandwidth in sub-6 GHz bands) that could yield bistatic range resolution below 2 m – sufficient to image vehicles and detect very small UAVs. The reference channel extraction from 5G NR signals is more complex than from broadcast waveforms because 5G uses beam-steered transmissions and dynamic resource allocation, requiring real-time decoding of the 5G physical layer to reconstruct the transmitted reference signal. Research demonstrations have shown promising results; operational PCL systems exploiting 5G are not yet widely fielded but represent the direction of capability development for the next decade.

Signal processing pipeline

The PCL signal processing chain is more computationally intensive per detection than conventional active radar, because the reference waveform is extracted from the environment rather than generated locally. The core stages are consistent across all PCL implementations, regardless of illuminator type.

Reference and surveillance channel acquisition. A PCL receiver requires at minimum two receive channels: a reference channel pointed at the illuminator to capture the direct-path signal as the processing reference, and one or more surveillance channels with antennas directed toward the surveillance volume. High-performance systems use arrays of 8–32 surveillance elements to provide spatial filtering capability for clutter rejection and direction-of-arrival estimation. The reference channel typically uses a directional antenna with high front-to-back ratio to maximize direct-path signal-to-noise ratio while rejecting target echoes.

Direct-path interference cancellation. The direct-path signal from the illuminator arrives at the surveillance channel 40–80 dB above target echo levels. Cancellation is performed by adaptive filtering: the reference channel signal is used to estimate and subtract the direct-path component from the surveillance channel. Algorithms such as least mean squares (LMS) or recursive least squares (RLS) are applied to track the slowly varying multipath interference. Cancellation depths of 50–80 dB are required; failure to achieve adequate DPI cancellation produces a noise floor that masks all but the strongest targets.

Cross-ambiguity function computation. After DPI cancellation, the surveillance channel signal is correlated against the reference channel across a range of bistatic delay (range) and Doppler frequency hypotheses. This two-dimensional cross-correlation – the cross-ambiguity function (CAF) – is the PCL equivalent of the matched filter in conventional radar. Each cell in the CAF corresponds to a specific bistatic range and radial velocity hypothesis. Target echoes appear as peaks in the CAF at their bistatic range and Doppler coordinates.

CAF computation is the dominant computational load in PCL processing. For a 1-second coherent processing interval at FM bandwidth, a single illuminator CAF requires on the order of 10⁹ multiply-accumulate operations. GPU acceleration using CUDA or OpenCL reduces this to sub-100 ms latency for real-time operation. High-channel-count systems processing multiple illuminators simultaneously require dedicated GPU compute nodes rather than general-purpose server hardware.

Clutter cancellation: ECA and STAP. Even after DPI cancellation, strong echoes from stationary terrain – hills, buildings, wind turbines – dominate the surveillance channel and must be suppressed before detection. The extended cancellation algorithm (ECA) applies a spatial filter across the surveillance array elements to project out the clutter subspace, exploiting the fact that clutter returns arrive from fixed azimuths and can be characterized from the data. Space-time adaptive processing (STAP) extends this to joint spatial-Doppler filtering, providing additional clutter discrimination for slow-moving targets. ECA-STAP implementations on modern GPU hardware achieve 40–60 dB clutter suppression.

CFAR detection and tracking. After clutter cancellation, a constant false alarm rate (CFAR) detector applies an adaptive threshold across the CAF to identify candidate target cells while maintaining a controlled false alarm rate independent of local noise and clutter levels. Detections are passed to the tracker, which applies Kalman filtering or multiple hypothesis tracking (MHT) to associate detections across processing intervals and form confirmed tracks. Track output is expressed in bistatic coordinates; conversion to Cartesian requires knowledge of the illuminator position and the receiver position – both of which must be surveyed to sub-100 m accuracy.

Multi-static fusion and air picture generation. A single PCL node provides bistatic range, Doppler, and – if using a surveillance array – direction-of-arrival. Two nodes sharing a common illuminator provide enough bistatic range pairs to reconstruct a Cartesian position. Three or more nodes or illuminators over-determine the position, enabling least-squares position estimation with accuracy typically in the 300–1,000 m range for FM-based systems and 50–200 m for DVB-T systems. Track outputs are formatted in ASTERIX Cat 48 or equivalent and fed to the SIGINT platform or air picture common operating picture.

Defense applications: what PCL enables

The zero-emission property of PCL creates operational capabilities that active radar cannot replicate. A PCL system deployed on a forward operating base reveals nothing about its presence to adversary electronic warfare assets. It provides persistent coverage without consuming the electromagnetic spectrum or creating an emitter signature that enemy targeting can exploit. In contested environments where active radar sites are routinely targeted, PCL can maintain air surveillance with no tactical signature.

A second and frequently underestimated advantage concerns stealth aircraft. Radar-absorbing materials are optimized for the microwave frequencies used by conventional fire-control and search radars (typically 3–18 GHz). FM and DAB PCL systems operate at VHF/UHF frequencies where the skin depth of radar-absorbing coatings is larger than the coating thickness and resonance effects in the aircraft structure can produce elevated bistatic RCS. Low-observable aircraft that appear essentially invisible to X-band active radar may produce detectable echoes on VHF PCL systems – a fact that has driven ongoing research into VHF PCL as a complement to conventional air defense radar.

PCL also provides a natural counter-drone surveillance layer when deployed with DVB-T illuminators. Small UAVs present an extremely challenging target for active radar – their low RCS, slow speed, and low altitude all work against conventional detection – but DVB-T PCL systems with adequate clutter cancellation have demonstrated repeatable detection of quad-rotor UAVs at 5–20 km, sufficient to provide alert and cueing for more focused RF sensors or kinetic interceptors. Integration with drone detection RF software cueing chains has been demonstrated in multiple European and Israeli field programs.

Integration with command and control

A PCL system's value is realized only when its track data reaches the operators and systems that can act on it. Integration with the air picture and the broader command and control system requires attention to both data format and latency.

Track data is typically formatted in ASTERIX (All-purpose Structured Eurocontrol Surveillance Information Exchange) Category 48 for monoradar tracks or Category 240 for sensor video. Systems feeding NATO common air picture infrastructure may alternatively use VMF (variable message format) or STANAG 4607 if ground moving target indicator integration is required alongside air tracks. Link 16 integration – passing PCL tracks into the tactical air picture – requires a waveform-capable terminal and is the standard for NATO base air defense installations.

Latency requirements for air defense cueing typically mandate end-to-end track latency below 2 seconds from target echo to display update. This drives the GPU-accelerated CAF computation described earlier and constrains the permissible coherent processing interval – longer integration improves detection sensitivity but increases latency. A 1-second coherent processing interval with 100 ms GPU processing and 500 ms tracking pipeline latency is achievable on current hardware and meets air defense latency requirements for most target classes.