Radar Link Budget Calculator | SNR & Integration

What Is a Radar Link Budget?

A radar link budget is the end-to-end accounting of every gain and loss between a radar’s transmitter and the point where the receiver decides whether a target is present. It answers a single question: does the signal-to-noise ratio (SNR) at the detector exceed the threshold needed for a specified probability of detection and false-alarm rate? Unlike a one-way communications link budget, the radar version must account for the round-trip path—out to the target and back—plus the target’s radar cross section (RCS), making path loss proportional to R⁴ rather than R².

The Radar Range Equation as a Budget

The standard form of the radar range equation, often called the radar equation, gives received SNR directly:

SNR = (P_t · G_t · G_r · λ² · σ) / ((4π)³ · R⁴ · k · T_s · B_n · L)

Each symbol maps to a line in the budget:

Working in decibels turns the multiplication and division into addition and subtraction, which is why engineers call it a “budget.” A positive bottom line means the target is detectable; the margin above the detection threshold tells you how robust the link is against fading, clutter, and component aging.

Walking Through Each Budget Line

Transmit Power and Antenna Gain

Peak transmit power (P_t) sets the energy illuminating the target. For a pulsed radar, the average power matters for search performance, so the duty cycle enters the budget when you convert peak to average. A 77 GHz automotive radar front-end may transmit only 10–13 dBm (10–20 mW), while a ground-based surveillance radar can push 60+ dBW. Antenna gain (G_t, G_r) depends on aperture area and efficiency; a 30 dBi antenna at 10 GHz has a physical aperture of roughly 0.07 m². When transmit and receive share the same antenna, G_t = G_r and you add the gain twice.

Wavelength and Target RCS

The λ² term is often absorbed into the antenna gain definition, but it appears explicitly in the range equation because gain is defined relative to an isotropic radiator. At 77 GHz (λ ≈ 3.9 mm), λ² is about −48 dBsm, a significant “loss” compared with a 3 GHz system (λ ≈ 100 mm, λ² ≈ −20 dBsm). Target RCS (σ) varies enormously: a pedestrian may present 0.5–1 m² (−3 to 0 dBsm), while a large truck can exceed 100 m² (+20 dBsm). Always budget for the smallest target you must detect at maximum range.

Noise Floor: kTB and Noise Figure

Thermal noise power equals k · T_s · B_n, where k is Boltzmann’s constant (−228.6 dBW/K/Hz), T_s is system noise temperature, and B_n is noise bandwidth. In practice, most engineers use receiver noise figure (NF) referenced to 290 K: noise power = −174 dBm/Hz + NF (dB) + 10 log₁₀(B_n). For a receiver with NF = 5 dB and B_n = 1 MHz, the noise floor is −174 + 5 + 60 = −109 dBm. The Friis noise-figure cascade tool helps you verify that LNA, mixer, and IF-chain contributions stay within your budget allocation.

System Losses

Losses that chip away at SNR include:

• Feed and transmission-line losses – coax, waveguide, or microstrip from transmitter to antenna. At 77 GHz on standard microstrip, loss can reach 1–2 dB/cm; careful material selection (e.g., RO3003 with Df ≈ 0.0013 at 10 GHz) is critical.
• Radome insertion loss – typically 0.5–1.5 dB for a well-designed radar radome, more if the radome is wet or iced.
• Atmospheric and rain attenuation – negligible below 10 GHz at short range, but 0.4 dB/km at 77 GHz in clear air and several dB/km in heavy rain.
• Processing losses – filter mismatch, A/D quantization, windowing; 1–3 dB is a common aggregate allowance.
• Scan loss – for phased arrays, element pattern roll-off at wide scan angles. At ±60° it can cost 3–4 dB.

Summing these carefully is where many budgets go wrong. A 1 dB underestimate in total loss translates directly into 1 dB less margin—or a 6 % reduction in maximum detection range (since range scales as the fourth root of SNR).

Radar Link Budget and Integration Gain Calculator

Enter your transmit power, antenna gains, frequency, target RCS, range, noise figure, bandwidth, and number of integrated pulses below. The calculator returns single-pulse SNR and the SNR after coherent or non-coherent integration, so you can see exactly how many pulses you need to close the budget.

For example, a 77 GHz automotive radar with P_t = 12 dBm, G_t = G_r = 20 dBi, σ = 1 m², R = 100 m, NF = 10 dB, B_n = 5 MHz, and 3 dB total system loss yields a single-pulse SNR of roughly 15 dB. If the detection threshold for P_d = 0.9 and P_fa = 10⁻⁶ is about 13.2 dB (Swerling I target), you have only 1.8 dB margin—thin for real-world variation. Coherently integrating 16 pulses adds ~12 dB, pushing the margin to a comfortable 13.8 dB and extending maximum range or allowing detection of smaller targets. For a deeper treatment of integration mathematics, see the radar SNR and integration gain discussion.

Pulse Integration: Closing the Budget Gap

Coherent vs. Non-Coherent Integration

Coherent integration sums return signals in phase, giving an ideal gain of 10 log₁₀(N) dB for N pulses. It requires accurate phase tracking across pulses, which is straightforward in FMCW radars but demands stable local oscillators and short coherent processing intervals (CPI). Non-coherent integration sums magnitudes (or squared magnitudes) after envelope detection, yielding a gain somewhat less than 10 log₁₀(N); the Albersheim approximation or Marcum–Swerling curves give the exact improvement for a given P_d and P_fa. As a rule of thumb, non-coherent integration of N pulses provides roughly 10 log₁₀(N) − loss_nc dB, where loss_nc ranges from 1–3 dB depending on N and target model.

Integration and Dwell Time

The number of pulses you can integrate is limited by dwell time—how long the beam stays on the target. For a mechanically scanned antenna rotating at 15 rpm with a 2° beamwidth, the dwell on a point target is about 22 ms. At a PRF of 1 kHz that gives 22 pulses per dwell. Phased arrays can steer electronically and allocate dwell adaptively, trading scan rate for integration depth on weak targets.

Timing and Skew Considerations on the PCB

In digital beamforming radars each receive channel must be sampled synchronously; clock-skew between ADC channels directly corrupts coherent integration gain. A skew of Δt introduces a phase error of 2πfΔt at the signal frequency, which for a 77 GHz IF at 10 MHz and 100 ps skew is only 0.36°—manageable—but at a 1 GHz IF the same 100 ps becomes 36°, destroying coherent summation. Careful stackup design and matched-length clock distribution are essential. Use the skew budget tool below to verify that your PCB trace-length tolerances keep phase error within your integration-loss allocation.

If the calculator shows that your clock tree contributes 50 ps of skew on a 6-layer Rogers/FR-4 hybrid stackup, and your budget allows 100 ps total, you still have margin for connector and cable variation. Should the number be tight, consider the PCB stackup thickness calculator to evaluate whether moving the clock layer closer to its reference plane reduces propagation-delay mismatch.

Common Mistakes in Radar Link Budgets

• Forgetting two-way losses. Feed loss appears twice—once on transmit, once on receive. A 0.5 dB feed loss costs 1.0 dB in the budget.
• Using peak RCS instead of median. A Swerling I target’s RCS is exponentially distributed; the median is 0.69 times the mean. Budget for a realistic percentile.
• Ignoring atmospheric loss at millimeter-wave. At 77 GHz, even 200 m of clear-air path adds ~0.08 dB—small, but rain at 16 mm/h adds ~2 dB over the same distance.
• Double-counting antenna efficiency. If your antenna gain already includes efficiency (as measured gain does), do not subtract efficiency again as a separate loss.
• Neglecting ADC and processing losses. Quantization noise from a 12-bit ADC adds roughly 0.3 dB; windowing for sidelobe control can add another 1–2 dB.

From Budget to PCB: Practical Impact

The link budget drives PCB material selection and layout constraints. If the budget shows only 2 dB of margin, you cannot afford 1.5 dB of transmission-line loss in the RF front-end. That pushes you toward low-loss laminates like Rogers RO3003 (Df ≈ 0.0013) or Isola Astra MT77 (Df ≈ 0.0017), substrate-integrated waveguide routing, or even antenna-in-package solutions. Phase-calibration accuracy feeds back into integration gain; the radar calibration and alignment guide covers how to set a realistic phase-error budget across the array.

Power-distribution networks for AESA transmit modules need low-loss, equal-phase splitting. A Wilkinson divider design tool lets you size the quarter-wave sections for your operating frequency and verify isolation between ports, which matters when dozens of channels share a common feed.

Quick-Reference: Radar Link Budget Checklist

1. Define detection requirements: P_d, P_fa, target RCS, and maximum range.
2. Calculate required single-pulse SNR from Albersheim’s equation or Swerling curves (see radar signal characteristics on Wikipedia).
3. Sum all gains: P_t (dBW) + G_t + G_r + 10 log₁₀(σ) + 20 log₁₀(λ).
4. Sum all losses: spreading constant + 40 log₁₀(R) + noise floor + system losses.
5. Subtract losses from gains → single-pulse SNR.
6. Add integration gain (coherent or non-coherent) for the available number of pulses.
7. Compare final SNR to the detection threshold. Margin ≥ 3 dB is a reasonable starting point; ≥ 6 dB if the environment is harsh.
8. Iterate: adjust power, antenna gain, waveform, or integration depth until the budget closes with acceptable margin.

Frequently Asked Questions

How does a radar link budget differ from a communication link budget?

A communication link budget accounts for one-way propagation (free-space loss ∝ R²), while a radar budget accounts for the round trip and target scattering, making free-space loss proportional to R⁴. The radar budget also includes target RCS as a variable “gain” term.

What SNR is needed to detect a target?

It depends on the required probability of detection (P_d) and false-alarm probability (P_fa). For P_d = 0.9 and P_fa = 10⁻⁶ against a Swerling I target, the single-pulse SNR threshold is approximately 13.2 dB. Non-fluctuating (Swerling 0) targets require about 13.0 dB under the same conditions.

Can I close a radar link budget by increasing integration alone?

Integration helps, but it is bounded by dwell time, target coherence time, and processing resources. Coherent integration beyond the target’s coherence interval yields diminishing returns because the target phase decorrelates. In practice, a combination of higher power-aperture product and moderate integration is more robust than relying on integration alone.

Where does PCB loss appear in the budget?

PCB transmission-line loss is part of the system-loss term (L). It appears twice—once in the transmit path and once in the receive path. At millimeter-wave frequencies this can be the single largest contributor to system loss, making laminate selection a link-budget decision, not just a manufacturing one.