Closed-Loop Tracking
Understanding Closed-Loop Tracking
Tracking a moving target with an antenna beam requires a servo control loop that senses pointing error and drives mechanical or electronic beam steering to reduce it. The most fundamental tracking technique, monopulse, was developed in the 1940s and remains the standard for precision tracking radars. It uses at least two antenna feed elements to form simultaneous sum (Σ) and difference (Δ) beam patterns. The sum pattern is the normal pencil beam with maximum gain on boresight. The difference pattern has a deep null on boresight with lobes of opposite phase on either side. The ratio Δ/Σ, called the error voltage, is a monotonic function of angular offset from boresight, providing an unambiguous direction and magnitude for the pointing error.
The tracking servo is typically a Type 2 control system (two integrators) that can track a target moving at constant angular velocity with zero steady-state error. The natural frequency ωn and damping ratio ζ determine the transient response: underdamped systems (ζ = 0.5 to 0.7) acquire targets faster but overshoot, while critically damped systems (ζ = 1.0) settle without overshoot but respond more slowly. For phased array radars, electronic beam steering replaces the mechanical servo, enabling tracking update rates of 1,000+ Hz with zero mechanical inertia, but the beam-pointing accuracy depends on phase shifter quantization (typically 5 to 6 bits, giving 0.1 to 0.2 degree pointing resolution at broadside).
Tracking Loop Equations
ε = Re[Δ/Σ] = km × θerror (for small angles)
Thermal Noise Angular Error (RMS):
σθ = θ3dB / (km × √(2 × SNR))
Servo Natural Frequency (Type 2):
ωn = √(Ka / J) ; fn = ωn / (2π)
Where km = monopulse slope (typically 1.5 to 1.8 per beamwidth), θ3dB = 3 dB beamwidth, SNR = single-pulse signal-to-noise ratio, Ka = servo acceleration constant, J = antenna moment of inertia. Example: 1° beamwidth, km = 1.6, SNR = 20 dB gives σθ = 0.022°.
Tracking Technique Comparison
| Technique | Angular Accuracy | Pulses per Estimate | Amplitude Sensitivity | Application |
|---|---|---|---|---|
| Amplitude monopulse | 0.01 to 0.05 θ3dB | 1 | None (ratio-based) | Fire control, satellite track |
| Phase monopulse | 0.01 to 0.03 θ3dB | 1 | None (phase comparison) | Precision instrumentation |
| Conical scan | 0.05 to 0.2 θ3dB | 4 to 8 | High (AM modulation) | Legacy fire control |
| Sequential lobing | 0.1 to 0.3 θ3dB | 2 to 4 | High | Simple tracking radars |
| Phased array digital BF | 0.005 to 0.02 θ3dB | 1 | None (digital processing) | AESA, multi-target track |
Frequently Asked Questions
How does monopulse achieve sub-beamwidth accuracy?
Monopulse forms simultaneous sum (Σ) and difference (Δ) beams. The ratio Δ/Σ is a monotonic function of angular offset, independent of target amplitude. A single pulse provides 0.01 to 0.1 beamwidth RMS accuracy. For a 1-degree beamwidth antenna, this is 0.01 to 0.1 degree (0.6 to 6 arcminutes). Unlike conical scan, monopulse is immune to amplitude scintillation since both channels are measured simultaneously.
What determines tracking loop bandwidth?
Bandwidth must follow target motion while rejecting noise. Required bandwidth scales as √(a/(R × θ3dB)), where a is target acceleration, R is range, and θ3dB is beamwidth. A 9g fighter at 50 km with 2-degree beamwidth needs about 1 Hz. Geostationary satellites need 0.01 to 0.1 Hz. Missile defense tracking 100g targets at 10 km needs 50 to 100 Hz.
What causes tracking errors?
Major sources: thermal noise (0.01 to 0.05 beamwidths at 20 dB SNR), target glint from multiple scattering centers (0.1 to 0.5 beamwidths for extended targets), multipath at low elevation (<2 degrees), servo lag (proportional to angular velocity / bandwidth), and structural/wind disturbances (0.01 to 0.05 degrees RMS). Total error is the RSS of all sources.