Phased-Array Trigger Systems

• Phased-array triggers are systems that coherently sum signals across multiple array elements using precise phase control to enhance detection sensitivity.
• They apply interferometric beamforming techniques in radio neutrino experiments to achieve lower signal-to-noise thresholds and improved detection efficiency.
• Emerging designs integrate hybrid architectures and nonreciprocal phase control, enabling advanced signal routing and flexible operational states across various frequency domains.

Searching arXiv for papers on phased-array triggers and related implementations. Phased-array trigger denotes a class of trigger and control mechanisms in which relative phase, delay, or modulation state across multiple array elements determines whether signals are coherently enhanced, whether a beam is formed in a selected direction, or whether the array enters distinct transmit and receive states. In radio neutrino instrumentation, the term most directly refers to an interferometric trigger that time-aligns and coherently sums antenna signals so that impulsive Askaryan radiation adds constructively while thermal noise adds incoherently (Avva et al., 2016). In other parts of the literature, the trigger parameter is the phase of low-frequency modulation signals in nonreciprocal phased-array antennas (Zang et al., 2019), or the pair of radio-frequency controls of an optical frequency comb that set the carrier-phase step of an ultrashort pulse train and thereby determine a free-space far-field interference pattern (Kato et al., 2024). The same phrase also appears alongside array-level image vetoes and timing-aligned analog camera triggers that are related in architecture but are not true beamforming systems (Barcelo et al., 2013).

1. Scope and operating principle

A phased array, in the standard electromagnetic sense, generates radiation patterns whose shape and direction can be electronically controlled by tailoring the amplitude and phase of the signals that feed each element of the array. A phased-array trigger applies that control principle at the decision stage: it does not wait for one element to exceed threshold independently, but instead uses the expected spatiotemporal coherence of a signal across the array to decide whether the array should fire, steer, or change state. In impulsive radio detection, the central scaling is that signal voltage adds as $N$ while uncorrelated noise adds as $\sqrt{N}$, so the beamformed sensitivity improves approximately as

$$\mathrm{SNR}_{\text{beam}} \propto \sqrt{N}\,\mathrm{SNR}_{\text{single}}.$$

That scaling is the basic reason phased triggering lowers threshold relative to single-channel logic (Avva et al., 2016).

Taken together, the cited literature suggests three recurrent meanings of the term. First, it may denote a low-threshold interferometric trigger, as in in-ice radio neutrino experiments. Second, it may denote a phase-controlled operational state, where a low-frequency modulation phase or a direction-dependent phase shifter selects different transmission and reception responses. Third, in optical systems it may denote the control input that establishes a wavefront, even when the mechanism is not an event trigger in the detector sense. A common misconception is therefore that every array-level trigger is a phased-array trigger. The Cherenkov-camera literature explicitly distinguishes timing-aligned analog summation and image-topology filtering from true coherent beamforming, even though the underlying electronics also use precise timing and distributed decision logic (Barcelo et al., 2013).

2. Low-threshold interferometric triggers for radio neutrino detection

The most direct realization of a phased-array trigger is the interferometric radio-neutrino trigger developed for Askaryan detection in ice. The ground-based interferometric phased-array concept treats a cluster of closely spaced, low-gain antennas as a small interferometer. If the expected arrival direction is known, or scanned over a set of directions, the voltages from the antennas are time-shifted and coherently summed. Measurements in an anechoic chamber and at Summit Station, Greenland found no significant correlation between adjacent antennas, even when as closely spaced as physically possible, validating the assumption that thermal noise remains effectively uncorrelated. In the same development program, a hardware zero-delay beam matched the ideal coherent sum to within about $15\%$, and trigger studies extrapolated that a 16-antenna phased array can reach 50% efficiency at SNR $\sim 1$ for a 100 Hz per-beam trigger, whereas a single antenna needs SNR $\sim 4$ for the same efficiency (Avva et al., 2016).

NuPhase operationalized that logic in deep Antarctic ice as a compact, FPGA-based trigger array at ARA5. The deployment used 10 Vpol antennas and 2 Hpol antennas, but after deployment 3 of the 10 Vpols were inoperable, so the beamforming trigger used 7 working Vpol channels. Signals were digitized at 1.5 GSa/s with 7-bit resolution, coherently summed into multiple beams, and thresholded using a beam-power proxy accumulated over 16 samples, corresponding to about 10.7 ns. The practical trigger used 15 trigger beams spanning roughly $-53^\circ$ to $+47^\circ$ in elevation. In situ attenuation scans showed 50% trigger efficiency at SNR $\approx 2.0$ at about 0.75 Hz per beam, compared with about SNR $\approx 3.7$ for the standard ARA5 combinatoric trigger. Hardware-level simulations, validated with in situ measurements, predicted a best-case far-field threshold of SNR $\sqrt{N}$0 and a trigger-level effective detector volume increase of about 1.8× at 10–100 PeV relative to the standard ARA trigger (Allison et al., 2018).

The A5 phased-array analysis extended the trigger concept into a hybrid station architecture. The A5 phased array used 7 VPol channels for the beamforming trigger, formed 15 predefined zenith directions, and triggered when the power in a beam exceeds a threshold power in a 10 ns window, while maintaining an 11 Hz global trigger rate across all beams. After the traditional ARA DAQ was lost due to a USB port failure in 2020, six ARA VPol channels were connected directly to the PA DAQ, creating the A5/PA hybrid system. That hybrid configuration combined low-threshold beamforming with a wider reconstruction geometry, yielding strong azimuth reconstruction, about 2× better zenith accuracy for high-SNR events, and improved background rejection variables derived from a 2D correlation map and a coherently summed waveform. The analysis used 504 days of data from 2020–2021 and was performed blindly with a 10% sample dataset reserved for cut development (Dasgupta, 2024).

3. Hybridization and next-generation neutrino trigger systems

Subsequent designs generalize the phased-array trigger into a hybrid architecture in which phasing is used for threshold reduction and sparse baselines are used for reconstruction. A representative concept targets 100 PeV tau neutrinos with compact phased arrays at 1 km prominence and sparse, autonomous high-gain antennas between stations. The fiducial trigger stations use 24 dual-polarized dipoles per station, operate in 30–80 MHz, and are separated by about 3 km along a ridge. The trigger threshold is assumed to be $\sqrt{N}$1 noise, applied to either the beamformed phased-array trigger or the autonomous reconstruction antennas. Relative to 4 km prominence, 1 km prominence increases the peak effective area by a factor of 10 at 100 PeV but reduces the effective area at the highest energies by about a factor of 2. Increasing the phased trigger from 10 antennas to 24 antennas increases the effective area at 100 PeV by a factor of 3. In the optimistic fiducial scenario, the instantaneous point-source sensitivity improves by about an order of magnitude over GRAND and BEACON at 100 PeV, while the addition of sparse reconstruction antennas grows the aperture by nearly a factor of two at the highest energies (Wissel et al., 2024).

ARA-Next pushes the same direction into DAQ design. Its central claim is not a completed beamformer but an RFSoC-based DAQ that can support more advanced triggers than the present hardwired ARA hierarchy. The proposed hardware is the Real Digital RFSoC 4x2, which can sample up to 5 GSPS and generate signals up to 9.85 GSPS. Because the board has 8 ADC inputs while a full ARA station has 16 channels, the design uses a channel-compression scheme based on upconversion, combination, digital filtering, and downconversion. The intended trigger logic includes 3 hits across all 16 channels, tagging events with a physical time sequence of hits, variable-length waveforms, directional reconstruction at trigger level, template matching for cosmic-ray pulses that point toward the ice surface, anthropogenic tagging, and a trigger for “n out of N channels with Double Pulses.” The program explicitly cites the success of phased-array triggering in lowering thresholds and improving detection efficiency as precedent for a more flexible multi-trigger architecture (Giri et al., 2024).

4. Nonreciprocal and full-duplex phased-array control

In nonreciprocal arrays, the trigger variable is not a beam-power threshold but a phase-controlled spatiotemporal state. Time-modulated nonreciprocal phased-array antennas employ resonant antenna elements loaded with varactors, with capacitances

$\sqrt{N}$2

The transmitted and received array factors differ by the sign of the modulation phase,

$\sqrt{N}$3

This sign flip is the essential nonreciprocal mechanism: the same physical array can beam in one direction when transmitting and in a different direction, or suppress that direction, when receiving. The measured microwave prototype showed isolation levels over 40 dB at desired directions with overall loss below 4 dB, while undesired harmonics were at least 30 dB below the desired signals and reflected nonlinear harmonics were at least 43 dB below the excitation (Zang et al., 2019).

A second route to nonreciprocal-beam behavior avoids time harmonics by using transistor-based nonreciprocal phase shifters. In that design, a linear array has different TX and RX phase progressions,

$\sqrt{N}$4

so the transmit and receive beam maxima occur at different angles. The demonstrated 2.4 GHz prototype used 8 patch elements arranged as a $\sqrt{N}$5 array with 4 nonreciprocal phase shifters. Measured results showed a TX beam at 52.33°, an RX beam at 127.7°, more than 15 dB isolation between TX and RX beams, more than 5 dB power gain compared to a conventional reciprocal phased array, and more than 7 dB amplification in each direction at 2.4 GHz with about 9% bandwidth. This literature treats the phase-control setting itself as the operational trigger for full-duplex beam asymmetry rather than as a detector threshold (Karimian et al., 2020).

5. Optical phased-array triggering and longitudinal-mode control

The optical literature introduces a distinct but technically precise meaning of phased-array trigger: the control signal that establishes the phase relation among optical emission points. A notable example is the free-space optical phased array built from a multipass cavity (MPC) and an optical frequency comb (OFC). Here the array is not a chip-based waveguide OPA but a set of multiple optical emission points generated by the MPC, while the OFC controls the relative carrier phase of the emitted ultrashort pulses across the full optical bandwidth. The comb is a phase-controlled mode-locked laser with repetition rate $\sqrt{N}$6 and carrier-envelope offset frequency $\sqrt{N}$7. The carrier-phase step between pulses is set by their ratio; the paper gives the explicit example

$\sqrt{N}$8

The cavity length is matched to approximately

$\sqrt{N}$9

so successive pulses emerge from different spatial positions with a fixed phase relation. In that sense, the comb is both the phase generator for the phased array and the stabilization reference that suppresses environmental drift (Kato et al., 2024).

The demonstrated setup used an Er-doped fiber comb laser centered at 1.55 µm with $$\mathrm{SNR}_{\text{beam}} \propto \sqrt{N}\,\mathrm{SNR}_{\text{single}}.$$0 MHz, a 1f–2f interferometer, an EDFA, and an MPC with 17 potential optical antennas, although only 3 and 5 were used in the demonstrations. Three-antenna operation formed an optical dot pattern and showed that locking $$\mathrm{SNR}_{\text{beam}} \propto \sqrt{N}\,\mathrm{SNR}_{\text{single}}.$$1 to the fluctuating cavity length made the interference nearly constant without any mechanical delay stage. Five-antenna operation produced a more localized pattern, and changing the comb frequency ratio shifted the dot position; examples were reported at frequency ratios 4.078 and 7.518, with about a 10× intensity contrast between the dot and background. Broadband operation was demonstrated from 1530 to 1610 nm without changing the comb control or MPC alignment, which the authors present as evidence that the method is genuinely broadband (Kato et al., 2024).

Related integrated optical phased arrays implement the control input differently. The end-fire silicon OPA on an SOI platform uses integrated resistive thermo-optic heaters to impose a phase ramp across waveguides terminated at the chip edge, enabling $$\mathrm{SNR}_{\text{beam}} \propto \sqrt{N}\,\mathrm{SNR}_{\text{single}}.$$2 spacing of 775 nm, a measured FWHM angular width of 17°, and a total steering range of 64° at 1.55 microns (Kossey et al., 2017). A separate compact OPA uses serial grating antennas so that each grating acts both as an antenna and as an optical power tap, eliminating directional couplers; for a 9×9 array it reports 98.4% total optical power utilization and a footprint of approximately 66 µm × 66 µm, with a 30% linear dimension reduction relative to the cited earlier design (Zhang et al., 2019). A plausible implication is that optical uses of “phased-array trigger” are often best understood as phase-programming mechanisms rather than detector triggers.

6. Related but non-equivalent array triggers

Several array-trigger systems are closely related to phased-array triggers in timing discipline and distributed control but are not true coherent beamformers. The CTA analog trigger architecture is explicit on this point: it combines analog signals from neighboring pixels and neighboring camera clusters in a structured, timing-sensitive way, but it does not coherently phase-align telescope waveforms across the array to form a beam. Its L0 Sum Trigger includes an analog delay line for each channel programmable from 0 to 5.5 ns in steps of 250 ps, and the system supports trigger regions of 2, 3, or 4 clusters, corresponding to 14, 21, or 28 pixels. The proposed system can handle trigger rates up to 10 MHz with negligible dead time, the L0 and L1 decision is fast enough to handle rates up to 100 MHz, and the full trigger latency was measured to be 190 ns total, with 20 ns from L0 and L1 and 170 ns from the L1 distribution (Barcelo et al., 2013).

The broader CTA trigger comparison reaches a related conclusion. Of the Majority trigger, Analog sum trigger, and Flexible binary / digital trigger with pattern analysis, the flexible digital family is the closest architectural analogue to a phased-array-like concept because it processes sampled space-time patterns in FPGA logic. Yet the performance result emphasized in that comparison is that direct summation of pulses, independent of their amplitudes, leads to lower energy thresholds than conventional majority schemes, especially for Doublets, while the most favorable near-threshold behavior came from the high-bandwidth chain using 2.6 ns Gaussian pulse FWHM and 1 GHz Flash-ADC (Shayduk et al., 2013).

Array-level trigger intelligence can also operate on reduced-resolution image geometry rather than phase coherence. The Distributed Intelligent Array Trigger (DIAT) exchanges compact trigger images among nearby telescopes and uses the Parallax width discriminator $$\mathrm{SNR}_{\text{beam}} \propto \sqrt{N}\,\mathrm{SNR}_{\text{single}}.$$3 to decide whether to veto events before camera readout. With a practical threshold of $$\mathrm{SNR}_{\text{beam}} \propto \sqrt{N}\,\mathrm{SNR}_{\text{single}}.$$4 m and a pass-through criterion $$\mathrm{SNR}_{\text{beam}} \propto \sqrt{N}\,\mathrm{SNR}_{\text{single}}.$$5, the DIAT retains more than 90% of genuine gamma events for offsets up to about 3.5°, reduces the array trigger rate by about a factor of 7, and yields about a factor of 4 increase in collection area at the lowest energies. This is an array-level pre-readout veto, not waveform beamforming (Dickinson et al., 2018).

The H.E.S.S. phase 2 second-level trigger is still further from beamforming: it is a monoscopic topological filter operating on a 2-bit pixel map from the 28-meter telescope camera. Its combined denoising, clustering, and center-of-gravity algorithm reduces the charged-particle rate by about a factor of 3 while preserving around 80% of low-energy photons after the full cut chain, with an average L2 latency of about 38 µs (Moudden et al., 2010). Conversely, the Sub-Millimeter Array processor is a true phased-array beamformer but not a trigger system: it applies programmable digital delay lines to 8 antennas with about 0.1 ns precision and spools the phased sum to a Mark 5b VLBI recorder, clarifying that phased processing and phased triggering are related but not identical categories (Nagpal, 2012).

The boundary cases are therefore important. A phased-array trigger, in the strict sense, requires coherent multi-element phase or delay control to make the trigger or operational decision. Analog sum triggers, topological vetoes, and distributed image triggers may share timing alignment, neighborhood logic, or array-level decision making, but they are only partially analogous unless they actually use phase-coherent summation to select sensitivity or state.