
What Is a Leaky Wave Antenna?
A leaky wave antenna (LWA) is a traveling-wave structure that radiates power continuously along its length through a controlled leakage mechanism. Unlike resonant antennas that radiate from a localized current distribution, an LWA produces a narrow beam whose pointing angle changes as a function of frequency—enabling electronic beam scanning without phase shifters. This makes leaky wave antennas attractive for radar sensors, point-to-point links, and imaging systems from roughly 10 GHz through 77 GHz and beyond.
The beam angle θ measured from broadside is governed by a simple dispersion relation: sin(θ) = β/k₀, where β is the phase constant of the guided mode and k₀ is the free-space wavenumber. When β is smaller than k₀ the structure radiates; when β equals k₀ the beam is at endfire; and when β passes through zero the beam crosses broadside. Engineering the β(f) curve is the central design task.
Uniform vs. Periodic Leaky Wave Antennas

Uniform (Continuous) LWA
A uniform LWA uses a waveguide with a continuous perturbation—an elongated slot in a substrate-integrated waveguide (SIW), for example—that leaks energy along its length. Because the fundamental mode is a fast wave (β < k₀) the structure radiates without any periodicity. Beam scanning is inherently one-sided: the beam sweeps from near-broadside toward endfire as frequency increases, but cannot cross broadside to the opposite side. A half-mode SIW leaky wave antenna on a 0.254 mm (10 mil) Rogers RO3003 laminate (Dk 3.00, Df 0.0013 at 77 GHz) is a common choice for automotive radar fan beams.
Periodic LWA
A periodic LWA introduces a repeating unit cell—interdigital slots, complementary split-ring resonators (CSRRs), or series-fed patch elements—spaced at period p. The periodicity produces space harmonics with phase constants β_n = β₀ + 2πn/p. The n = −1 harmonic is normally the radiating one, and its β can sweep through zero, allowing the beam to scan continuously from backward through broadside to forward. This full-space scanning capability is the chief advantage of periodic structures, though a gain dip at broadside (the “open stopband”) must be suppressed by careful unit-cell design.
| Feature | Uniform LWA | Periodic LWA |
|---|---|---|
| Radiating mechanism | Fast-wave leakage of fundamental mode | Space-harmonic (n = −1) radiation |
| Beam scan range | Forward quadrant only | Backward through broadside to forward |
| Broadside radiation | Not inherent | Possible with stopband suppression |
| Typical PCB form | Slotted SIW, half-mode SIW | CRLH microstrip, series-fed patches, meander |
| Design complexity | Lower | Higher (unit-cell optimization) |
Dispersion Engineering: The CRLH Approach
The composite right/left-handed (CRLH) transmission line is the most widely used platform for periodic leaky wave antennas on PCB. A CRLH unit cell combines a series capacitance (interdigital fingers or a gap) with a shunt inductance (a grounded stub or via) alongside the conventional series inductance and shunt capacitance of the host line. When the series and shunt resonances are balanced—ω_se = ω_sh—the dispersion curve passes smoothly through β = 0 with no stopband, enabling seamless broadside radiation.
Designing a CRLH LWA on, say, a 0.508 mm (20 mil) Rogers RO4350B substrate (Dk 3.48, Df 0.0037) at a center frequency of 24 GHz proceeds roughly as follows:
- Choose the unit-cell period p. Keep p below roughly λ_g/4 to stay in the effective-medium regime; at 24 GHz on RO4350B that means p ≤ 1.6 mm.
- Size the series capacitor. Interdigital fingers or a lumped gap set C_L; target the balanced condition ω₀ = 1/√(L_R · C_L) = 1/√(L_L · C_R).
- Size the shunt inductor. A short via-grounded stub sets L_L. Via diameter and anti-pad clearance affect parasitic capacitance at mmWave.
- Full-wave tune. Run an eigenmode simulation (HFSS, CST) of a single unit cell with periodic boundaries to extract the Bloch dispersion diagram. Adjust geometry until the curve passes through β = 0 at the desired center frequency with no gap.
- Set the antenna length. More cells give a narrower beam but increase loss. A 20-cell CRLH line at 24 GHz typically yields a 5°–8° half-power beamwidth and scans ±40° over a 4 GHz bandwidth.
Understanding the differences between high-frequency and high-speed design is essential when routing the feed network, because leaky wave structures are sensitive to both phase velocity and attenuation constant.
PCB Substrate and Stackup Considerations
Substrate choice directly controls the leakage rate α, the phase constant β, and the achievable scan range. Key parameters:
- Dielectric constant (Dk). Lower Dk widens the fast-wave region, increasing the scan range for a uniform LWA. PTFE-based laminates like RO3003 (Dk 3.00) or RT/duroid 5880 (Dk 2.20) are preferred above 40 GHz.
- Loss tangent (Df). Df adds to the total attenuation α_total = α_rad + α_d + α_c. If dielectric loss dominates, aperture efficiency drops because power is absorbed before it can radiate. At 77 GHz, moving from FR-4 (Df ~ 0.02) to RO3003 (Df 0.0013) can recover 3–5 dB of gain.
- Substrate thickness. Thicker substrates increase the leakage rate of slot-based LWAs, shortening the effective aperture. For a 24 GHz SIW LWA, 0.508 mm is a practical starting point; for 77 GHz, 0.127–0.254 mm keeps surface-wave excitation under control.
- Copper roughness. At millimeter-wave frequencies, conductor loss is significant. Specify rolled or very-low-profile (VLP) copper foil (Rz ≤ 1.5 µm) on the signal layer.
Material datasheets from Rogers RO4000 series laminates provide Dk/Df versus frequency curves essential for accurate dispersion modeling. When scaling a design across frequency bands, the guidance in our article on RF design frequency scaling applies directly to leaky wave unit cells.
Beam Scanning Performance and Trade-Offs
Scan Rate and Bandwidth
The scan sensitivity dθ/df depends on the slope of the dispersion curve. A CRLH LWA on RO4350B at 24 GHz typically scans at roughly 10°/GHz, so a 2 GHz sweep covers ±10° around broadside. Steeper dispersion (higher group velocity near β = 0) gives faster scanning but narrows the instantaneous bandwidth for a fixed beam direction—a fundamental trade-off.
Gain and Beamwidth
Directivity in the scan plane scales with the effective aperture length L_eff, which is limited by the total attenuation. A rule of thumb: choose L so that 90 % of input power radiates (α_rad · L ≈ 2.3 Np). For a 20-cell CRLH LWA with α_rad ≈ 0.1 Np/cell, that means roughly 23 cells. Cross-plane beamwidth is set by the antenna’s transverse dimension or by arraying multiple LWA lines in parallel.
Open Stopband Suppression
At broadside (β = 0) a periodic LWA can exhibit a gain null because contra-directional coupling creates a stopband. Proven suppression techniques include asymmetric unit cells, stub-loaded CRLH cells, and the use of metasurface-based unit cells that break the symmetry of the Bloch impedance.
Practical Layout and Fabrication Guidelines

Feed and Termination
An LWA is a two-port device. The input port is impedance-matched (typically 50 Ω); the far end is terminated in a matched load to absorb unradiated power and suppress standing-wave ripple on the pattern. At 24 GHz a 0402 thin-film resistor on the output microstrip works; at 77 GHz an on-board tapered resistive termination is more repeatable.
Via Fencing for SIW Structures
SIW-based LWAs require rows of plated vias to form the waveguide sidewalls. Via pitch should be below λ_g/5 and via diameter at least half the pitch to prevent leakage through the fence. For a 24 GHz SIW on RO4350B (λ_g ≈ 6.7 mm), that means via pitch ≤ 1.3 mm and via diameter ≥ 0.6 mm—well within standard PCB drill capability.
Tolerance Sensitivity
Leaky wave antennas are more sensitive to etching tolerance than patch arrays because the leakage rate depends on small geometric perturbations. A ±25 µm (±1 mil) etch variation on a 0.15 mm interdigital gap at 24 GHz shifts the balanced frequency by roughly 200 MHz and distorts the beam at broadside. Specify IPC-6012 Class 3 or tighter etch tolerance, and read our guide on moving antenna prototypes to production for yield strategies.
Assembly Considerations for Hybrid LWA Modules
When a leaky wave antenna board carries active components—LNAs, mixers, or power amplifiers for an active electronically scanned array (AESA)—mixed assembly processes come into play. Surface-mount devices on the non-radiating side are commonly wave-soldered, which requires adhesive pre-application and selective pallets.
SMD Adhesive Estimation for Wave Soldering
Before wave soldering bottom-side passives on an LWA panel, estimate the adhesive volume needed per board. Enter the number of SMD pads, average dot diameter, and panel quantity below to get a total adhesive weight.
For a typical 24 GHz LWA module with 60 bottom-side 0402 passives (dot diameter ~0.6 mm), the calculator returns roughly 0.15 g of adhesive per board. Under-application causes tombstoning during the wave; over-application contaminates pads and degrades RF grounding—both critical failure modes on a leaky wave feed network.
Wave Solder Pallet Planning
Selective wave soldering of through-hole connectors (SMA, SMPM) on an LWA board requires a pallet that exposes only the target joints while protecting the radiating aperture and fine-pitch components. Use the calculator below to estimate pallet requirements based on board dimensions and production volume.
A 100 × 40 mm LWA panel running 500 units per week typically needs 4–6 pallets in rotation to allow cleaning cycles. Pallet flatness should be held to ±0.2 mm to prevent solder bridging on the dense via fences of an SIW structure.
Measurement and Validation
Characterizing a leaky wave antenna requires both impedance and radiation measurements. S₁₁ and S₂₁ of the two-port structure are measured first; S₂₁ quantifies how much power reaches the termination (ideally below −10 dB for high radiation efficiency). Radiation patterns must be swept across the operating bandwidth to verify the scan curve. A compact antenna test range or planar near-field scanner is appropriate; details on setup are covered in our antenna measurement setup guide.
Key measurements to capture:
- Beam angle vs. frequency (compare against the simulated dispersion curve).
- Gain vs. frequency—watch for a dip at broadside if the open stopband is not fully suppressed.
- Cross-polarization level, especially for SIW slot LWAs that can excite unwanted modes.
- Realized gain vs. scan angle, accounting for the cos(θ) projected-aperture roll-off.
For far-field pattern details including gain, beamwidth, and null placement, the antenna pattern measurement article provides complementary guidance.
Applications of Leaky Wave Antennas
Automotive Radar (76–81 GHz)
Half-mode SIW LWAs produce a narrow elevation fan beam that scans with the FMCW chirp bandwidth, enabling angle-of-arrival estimation without a full phased-array beamformer. This reduces cost and DC power in short-range corner radars.
5G Fixed Wireless and Backhaul
At 24–28 GHz, CRLH LWAs can serve as frequency-scanned sector antennas covering ±30° in azimuth. Combined with a transmitarray lens in elevation, they form a low-cost two-dimensional scanning solution.
Security Imaging
Passive millimeter-wave imagers at 35 GHz and 94 GHz use LWA line sources scanned mechanically in one axis and by frequency in the other, achieving real-time imaging frame rates without expensive T/R modules.
Satellite Ground Terminals
Flat-panel Ku/Ka-band terminals for LEO constellations are exploring LWA-based sub-arrays to reduce the number of phase-shifter channels. The inherent frequency scanning supplements electronic beam steering for elevation tracking. Further discussion of flat-panel architectures is available in our satellite antenna PCB design overview.
FAQ
How does a leaky wave antenna scan its beam?
The beam angle is set by the ratio of the guided phase constant β to the free-space wavenumber k₀: sin(θ) = β/k₀. Because β varies with frequency, changing the operating frequency steers the beam without any mechanical movement or electronic phase shifters.
Can a leaky wave antenna radiate at broadside?
A periodic LWA can radiate at broadside when its phase constant passes through zero. However, an open stopband often causes a gain dip at that point. Balanced CRLH unit cells or asymmetric periodic cells suppress this stopband and enable continuous broadside radiation.
What substrate is best for a millimeter-wave LWA?
Low-loss PTFE-ceramic laminates such as Rogers RO3003 (Dk 3.00, Df 0.0013) or RT/duroid 5880 (Dk 2.20, Df 0.0009) are standard choices above 60 GHz. They minimize dielectric attenuation, preserving aperture efficiency over long radiating lengths.
How long should a leaky wave antenna be?
Length is chosen so that approximately 90 % of input power is radiated: α_rad · L ≈ 2.3 nepers. A longer antenna narrows the beam but increases ohmic loss. Typical lengths range from 5 λ₀ to 15 λ₀ depending on the acceptable beamwidth and loss budget.
What is the difference between a leaky wave antenna and a series-fed patch array?
A series-fed patch array radiates from discrete resonant elements and is designed to suppress inter-element radiation, while an LWA radiates continuously (or quasi-continuously) along its length. Both exhibit frequency-dependent beam squint, but an LWA exploits it intentionally for beam scanning, whereas in a series-fed array it is usually an unwanted side effect. For related broadband approaches, see the IEEE/Wikipedia overview of leaky-wave antennas.
Ready for a quotation?
Send your requirements for a material, stackup and manufacturing review. Price and schedule are confirmed for your project.
