UAV Radar Design: Airborne Sensing Basics

UAV radar design requires balancing size, weight, power, and cost (SWaP-C) against the need for sufficient range, resolution, and Doppler sensitivity to detect and classify ground or air targets from a moving platform. Developers must understand how platform motion, antenna aperture, and waveform choices affect performance before selecting components. This article explains the core principles, key design trade-offs, and common pitfalls in airborne radar for drones.

Formula

The key relationships are the Doppler shift and the unambiguous range/velocity limits. For a monostatic radar (transmitter and receiver co-located), the Doppler frequency shift f_d is given by:

f_d = (2 * v_r * f_c) / c

where v_r is the radial velocity of the target relative to the radar (m/s), f_c is the carrier frequency (Hz), and c is the speed of light (≈ 3 × 10⁸ m/s). The maximum unambiguous Doppler velocity is v_max = (PRF * c) / (4 * f_c), where PRF is the pulse repetition frequency. The maximum unambiguous range is R_max = c / (2 * PRF).

Worked Example

Using the default values: altitude = 120 m, platform speed = 15 m/s, frequency = 10 GHz, PRF = 2000 Hz.

• Maximum unambiguous Doppler velocity: (2000 * 0.03) / (2 * 10) = 3.00 m/s. This means a target moving faster than 3 m/s radially will be aliased (folded) in Doppler.
• Maximum unambiguous range: (0.03 * 1,000,000) / (2 * 2000) = 7.5 km. This is the range beyond which echoes from a second pulse arrive after the next pulse is sent, causing range ambiguity.
• Doppler shift for a target with the same speed as the platform (15 m/s radial): (2 * 15 * 10) / 0.03 = 10,000 Hz. This is the shift you would see from a stationary ground target directly ahead (closing).

The calculator shows that a 10 GHz radar with PRF 2000 Hz cannot unambiguously measure radial velocities above 3 m/s, which is far below the platform speed of 15 m/s. This is the classic problem of medium-PRF vs. high-PRF design.

Background: Why UAV Radar is Different

Radar on a drone faces constraints that ground-based or large airborne radars do not. The platform is small, so the antenna aperture is limited, which reduces angular resolution and gain. Power is scarce, limiting range. Weight and size restrictions force the use of lightweight materials and compact electronics, often at the expense of processing power. Additionally, the radar moves, so ground clutter is Doppler-shifted, and the radar must distinguish between returns from the ground and returns from actual moving targets. This is the fundamental challenge of moving target indication (MTI) and ground moving target indication (GMTI).

Developers must decide what the radar is for: detecting large ground vehicles, mapping terrain, avoiding obstacles, or tracking small drones. Each mission changes the design priorities. A terrain-mapping radar needs high range resolution and a wide swath, while a target-detection radar needs high Doppler resolution and low false alarm rate.

How Airborne Sensing Works

An airborne radar transmits radio waves and listens for echoes. The time delay gives range, the Doppler shift gives radial velocity, and the antenna beam direction gives angle. In a moving platform, the ground produces a broad Doppler spectrum because different parts of the ground have different relative velocities. The radar must filter out this clutter to see moving targets. This is done using pulse-Doppler processing, where multiple pulses are coherently integrated to produce a Doppler spectrum.

Key parameters include:

• Waveform: simple pulse, linear frequency modulation (LFM) for pulse compression, or phase-coded waveforms.
• Pulse repetition frequency (PRF): low PRF gives long unambiguous range but ambiguous Doppler; high PRF gives unambiguous Doppler but ambiguous range.
• Antenna: phased array or mechanically scanned. Phased arrays allow electronic beam steering, which is faster and more reliable for a moving platform.
• Processing: fast Fourier transform (FFT) over pulses to generate Doppler bins, followed by constant false alarm rate (CFAR) detection.

For a UAV, the antenna is often a small array of patches or a slotted waveguide. The beamwidth is inversely proportional to the aperture size in wavelengths. At 10 GHz (X-band), a 0.3 m antenna gives a beamwidth of about 5.7 degrees. At 35 GHz (Ka-band), the same size gives about 1.6 degrees, but range is reduced due to higher atmospheric attenuation.

How to Choose the Right Design

Choosing a radar design starts with the mission. Define the minimum detectable target radar cross-section (RCS), the required detection range, the range and Doppler resolution, and the update rate. Then consider the platform constraints.

1. Determine the frequency band: L-band (1-2 GHz) for long range and foliage penetration, S-band (2-4 GHz) for weather, C-band (4-8 GHz) for general purpose, X-band (8-12 GHz) for high resolution and small antennas, Ku/Ka-band (12-40 GHz) for very high resolution but short range.
2. Calculate the required power-aperture product: the product of average transmit power and antenna aperture determines the signal-to-noise ratio (SNR) for a given range and RCS. Use the radar equation.
3. Select the waveform and PRF. For ground moving target indication, a medium PRF (e.g., 10-20 kHz) is often used to balance range and Doppler ambiguities, but requires complex ambiguity resolution.
4. Choose the antenna type: a phased array is preferred for agile beam steering, but a simpler mechanically scanned antenna may suffice for a fixed forward-looking radar.
5. Plan the processing chain: real-time FFT and CFAR require a GPU or FPGA with sufficient throughput. The processing load scales with the number of range bins, Doppler bins, and antenna beams.

For example, a small drone detecting a car (RCS ~10 m²) at 1 km with a 10 GHz radar needs about 1 W average power and a 0.2 m aperture to achieve a SNR of 10 dB, assuming a 1 Hz update rate. But if the drone is moving at 20 m/s, the clutter spectrum is wide, and you need a high PRF to avoid Doppler ambiguity.

Common Problems and Misconceptions

One common mistake is ignoring platform motion compensation. The radar’s own velocity must be precisely known to remove the Doppler shift of ground clutter. This requires an inertial navigation system (INS) and GPS, and errors in velocity cause clutter leakage that masks slow targets.

Another issue is range-Doppler ambiguity. Many developers try to use a low PRF to get long unambiguous range, but then the Doppler is ambiguous, and moving targets appear at wrong velocities. Conversely, a high PRF gives unambiguous Doppler but range ambiguity. The solution is to use multiple PRFs and resolve ambiguities, or to use a waveform with a high duty cycle such as FMCW (frequency-modulated continuous wave). FMCW is popular for small UAVs because it uses low peak power and provides high range resolution, but it has limited dynamic range and is susceptible to self-interference.

A common misconception is that higher frequency is always better. While higher frequencies give better resolution for the same aperture, they suffer from higher atmospheric attenuation and are more affected by rain. At 94 GHz, attenuation is about 0.4 dB/km in clear air, but in rain it can be 10 dB/km or more. For a small UAV with limited power, this can reduce range drastically.

Cost and Size Trade-offs

Cost is a major driver. A full phased array with hundreds of T/R modules can cost tens of thousands of dollars and weigh several kilograms. For a small drone, a simpler design using a single channel and a mechanically scanned antenna may cost under $1,000 and weigh less than 500 g. However, such a radar has a slow scan rate and cannot track multiple targets simultaneously.

Alternatively, a frequency-modulated continuous wave (FMCW) radar using a single chip such as those used in automotive applications can be very cheap and small. These operate at 77 GHz and have integrated antennas, but their range is limited to a few hundred meters and they are not designed for airborne use. Developers often adapt automotive radar chips for drones, but must account for the different clutter environment and the need for higher Doppler resolution.

Power consumption is also critical. A typical small UAV battery provides about 100 Wh. A radar that consumes 20 W can run for 5 hours, but that reduces flight time. Many developers use duty cycling: the radar only transmits for short bursts, reducing average power.

Safety and Regulatory Considerations

Radar emissions are regulated by national authorities. In the US, the Federal Communications Commission (FCC) regulates the use of radio frequencies. UAV radars typically operate in bands designated for radar, such as 9.3-9.5 GHz (X-band) or 76-81 GHz (automotive). Developers must ensure their radar does not interfere with other users. Also, the radar must not be a hazard to people: the specific absorption rate (SAR) for human exposure is limited, but for a UAV flying at altitude, the risk is low.

Another safety aspect is the radar’s effect on other aircraft. A UAV radar transmitting high power could interfere with other aircraft’s radar or communication systems. Therefore, many UAV radars are designed to operate in a low-power mode when near populated areas. Also, the radar must be reliable; a failure could cause the drone to crash, so redundant sensors are often used.

Alternatives to Traditional Radar

For some UAV applications, other sensing modalities may be more appropriate. LiDAR (light detection and ranging) provides high-resolution 3D mapping but is sensitive to weather and has short range. Cameras with computer vision are cheap and lightweight but have poor performance in low light and cannot measure velocity directly. Ultrasonic sensors are very short-range and only useful for obstacle avoidance at low speeds.

Radar has the advantage of working in all weather, day or night, and providing direct velocity measurement via Doppler. For ground moving target indication, radar is the only practical option. However, for terrain mapping, LiDAR often gives better resolution. Many UAVs use a combination: radar for long-range detection and tracking, and LiDAR or cameras for close-in obstacle avoidance.

Frequently Asked Questions

What is the difference between FMCW and pulse-Doppler radar for UAVs?

FMCW radar transmits a continuous wave with a frequency ramp, and range is determined by the beat frequency between the transmitted and received signals. It uses low peak power and is simple to implement, making it popular for small UAVs. Pulse-Doppler radar transmits short pulses and uses time delay for range and Doppler shift for velocity. It can achieve longer range and better clutter rejection but requires higher peak power and more complex processing.

How does platform motion affect the radar’s ability to detect slow-moving targets?

Platform motion causes the ground clutter to have a Doppler spread. If the platform moves at 15 m/s, the clutter from directly ahead is shifted by about 10 kHz at 10 GHz. A slow-moving target (e.g., a walking person at 1 m/s) has a Doppler shift of only about 667 Hz. To separate the target from clutter, the radar must have a Doppler resolution better than the difference, and the clutter must be suppressed. This is done using high-pass filtering (MTI) or space-time adaptive processing (STAP).

What is the maximum range of a small UAV radar?

The maximum range depends on transmit power, antenna gain, frequency, and target RCS. For a typical small UAV radar with 1 W average power, a 10 GHz frequency, and a 0.2 m antenna, the range for a 1 m² target (like a small drone) is about 1-2 km with a 10 dB SNR. For a large vehicle (10 m²), the range can be 3-5 km. However, these are optimistic estimates; in practice, clutter and atmospheric losses reduce range.

Can I use a commercial automotive radar chip for my UAV?

Yes, many developers use 77 GHz automotive radar chips because they are cheap and compact. However, they are designed for ground use, with a limited unambiguous range (about 200 m) and a narrow field of view. For airborne use, you need to modify the waveform and processing to handle the higher platform speed and the different clutter environment. The Doppler ambiguity can be a problem, as the chip’s PRF is often too low for airborne speeds.

Do I need a phased array antenna for a UAV radar?

Not necessarily. A phased array allows electronic beam steering, which is fast and reliable, but it is expensive and heavy. For a fixed forward-looking radar, a simple horn or patch antenna with a wide beam may suffice. For tracking multiple targets, you need a phased array or a mechanically scanned antenna. The choice depends on the mission: if you only need to detect targets in a narrow forward sector, a fixed antenna is enough.

What is the best frequency band for a small UAV radar?

For small UAVs, X-band (8-12 GHz) is a good compromise: it offers good resolution with a small antenna, moderate atmospheric attenuation, and a wide selection of components. Ku-band (12-18 GHz) and Ka-band (26-40 GHz) give better resolution but shorter range. L-band (1-2 GHz) is useful for foliage penetration but requires a large antenna. The best choice depends on your specific range and resolution requirements.

Related

• Core Keeper Server Default Port & How to Change It
• How to Capture BIOS/UEFI and Pre-Windows Boot Screens in Windows 11
• Google Table Tennis Robot: Amateur-Level Rally Play?

in t | Platform altitude | m | 120 | 0
in v | Platform speed | m/s | 15 | 0
in f | Radar frequency | GHz | 10 | 0
in prf | PRF | Hz | 2000 | 0
out vmax | Max unambiguous Doppler (m/s) | m/s | (prf * 0.03) / (2 * f) | 2
out rmax | Max unambiguous range (m) | m | (0.03 * 1000000) / (2 * prf) | 1
out fdopp | Doppler shift for target (m/s) | Hz | (2 * v * f) / 0.03 | 2