6. Specific Antenna Types

Beyond the basic dipoles and arrays, certain antenna geometries have become standardized for their unique properties.

6.1 The Yagi-Uda antenna

Invented by Hidetsugu Yagi and Shintaro Uda in 1926, the Yagi-Uda is the classic outdoor TV antenna and the workhorse of amateur VHF/UHF directional gain. It has one driven element (a half-wave dipole or folded dipole), one reflector behind it (slightly longer than $\lambda/2$, typically about 5% longer), and one or more directors in front (slightly shorter than $\lambda/2$, typically 5% shorter, getting shorter as you move forward).

   Reflector(longer)  Driven  Director1  Director2  Director3
        |              ●●        |          |          |
   spacing ~0.25λ   feed   ~0.15λ      ~0.20λ     ~0.25λ
        ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ (boom)

Why does it work? The driven element is fed; the reflector and directors are parasitic, with no feed, but they have currents induced by the driven element. The reflector is slightly inductive (longer than resonance) so its induced current lags; combined with the geometry, this makes a backward-pointing field cancel and a forward-pointing field reinforce. The directors are slightly capacitive (shorter than resonance) so their induced currents lead, further reinforcing the forward beam.

A 3-element Yagi (reflector, driven, director) gives about 7 dBi. Each additional director adds about 1 dB up to about 10 elements, then diminishing returns. A 15-element Yagi gives about 15 dBi.

Used in: outdoor TV antennas, ham radio VHF/UHF beams, point-to-point links, RFID readers, drone direction-finding. Very narrow band (a few percent), sensitive to nearby metal, but cheap and effective.

For a TEMPEST attacker, a hand-built Yagi at, say, 1 GHz can give 12 dBi of gain pointed at a target window, dramatically improving the chance of recovering a video signal radiating from a victim's monitor. The Yagi is the "telescope" of RF surveillance.

6.2 The log-periodic dipole array (LPDA)

A log-periodic antenna looks like a Yagi but with all the elements driven, and with element lengths and spacings that scale geometrically (each is a fixed ratio $\tau$ of the previous one). The driving phase alternates between elements (often via a crossed-feed boom).

The result: at any given frequency in a wide band, only a few elements are near resonance and they do most of the radiating. As frequency changes, the "active region" slides along the boom. This gives nearly constant gain (around 6–9 dBi) over a 4:1 to 10:1 frequency range.

Used in: outdoor "all-band" TV antennas (54 to 800 MHz), EMC test antennas, broadband amateur antennas, military and intelligence wide-band scanning. Whenever you need one antenna to cover a decade of bandwidth at modest gain, log-periodic is the answer.

6.3 The helical antenna

A helical antenna is wire wound in a helix above a ground plane.

• Normal mode (helix small compared to $\lambda$): radiates like a vertical dipole, broadside, linearly polarized. Used as compact replacement for whips on handheld radios.
• Axial mode (helix circumference $\approx \lambda$, pitch about 12–14°): radiates a circularly polarized beam along the helix axis. The CP comes naturally from the geometry: a wave traveling along a helical path is circularly polarized.

Axial-mode helices give about 12–18 dBi of gain depending on length, with broad bandwidth and natural circular polarization. Used in: GPS satellite transmit antennas (RHCP), Iridium satellite antennas, deep-space probes, command links to satellites where polarization mismatch from a tumbling spacecraft would otherwise kill the link.

The fact that GPS uses RHCP is why a left-hand-circular antenna near a GPS receiver is invisible to it: a 30 dB cross-polarization rejection makes it a useful trick for testing and even for some types of selective jamming.

6.4 Horn antennas

A horn is a flared open end of a waveguide. The waveguide is a hollow metal pipe carrying the wave (we cover waveguides in detail in Chapter 18). The flare gradually expands the waveguide cross-section to match the impedance of free space, minimizing reflection.

• E-plane sectoral horn: flared in the plane of the E-field only.
• H-plane sectoral horn: flared in the plane of the H-field only.
• Pyramidal horn: flared in both planes (rectangular cross-section).
• Conical horn: flared from a circular waveguide.

Gain ranges from 10 dBi for a small horn to about 25 dBi for a large pyramidal. Bandwidth is wide (multiple octaves possible). Pattern is clean with low side lobes.

Used in: feeds for parabolic dishes (the horn sits at the focal point and illuminates the dish), reference antennas in EMC measurements, satellite earth-station feeds, military radar feeds, microwave point-to-point links. The "calibrated standard gain horn" is the gold reference for antenna gain measurement, because its gain is calculable from its dimensions.

      Pyramidal horn
 
   waveguide       flared  
    input          aperture
       ┌─────────────────────┐
       │   ╱             ╲   │
       │  ╱               ╲  │
       │ ╱                 ╲ │
       │╱                   ╲│
       ─────●──────●─────────  → free space
       │╲                   ╱│
       │ ╲                 ╱ │
       │  ╲               ╱  │
       │   ╲             ╱   │
       └─────────────────────┘

6.5 Parabolic reflector (the "dish")

A parabolic reflector exploits the geometric property of a parabola: every ray from the focus, reflected off the parabolic surface, emerges parallel to the axis. So if you put an antenna (a feed) at the focus, the dish "collimates" the spherical wave from the feed into a flat plane wave aimed along the axis. Receiver-side: a plane wave from far away is focused onto the feed.

The aperture gain is

$G \approx \eta_{ap} \cdot \frac{4\pi A_{phys}}{\lambda^2}$

with $\eta_{ap}$ around 0.55 to 0.7. The HPBW is approximately

$\text{HPBW} \approx 70 \cdot \frac{\lambda}{D} \text{ degrees}$

For a 1 m dish at 10 GHz: $A_e \approx 0.6 \cdot \pi (0.5)^2 = 0.47$ m², so $G \approx 4\pi \cdot 0.47 / (0.03)^2 \approx 6500 \approx 38$ dBi. HPBW $\approx 70 \cdot 0.03 / 1 = 2.1°$. A pencil beam.

The dish itself is passive (no power gain in the active sense; you cannot get out more than you put in). But it acts as a passive amplifier in directivity terms: it doesn't generate energy, but it concentrates it from a narrow incoming or outgoing solid angle. Newcomers sometimes wonder how a dish "amplifies" RF; it is the same way a magnifying glass focuses sunlight to a hot spot. No new energy, just spatial concentration.

Variants:

• Prime focus: feed at the focus, blocking some of the aperture. Simple, slightly lossy due to blockage.
• Cassegrain feed: a small subreflector intercepts the feed's waves and directs them back to a feed at the dish's vertex. This avoids long feed lines from feed back to receiver and reduces blockage. Used in NASA's Deep Space Network and in deep-space planetary radars.
• Gregorian feed: like Cassegrain but with an ellipsoidal subreflector, slightly different optical properties.
• Offset feed: the parabolic surface is a slice of the parabola off-axis, with the feed offset to one side. No blockage, and the dish can be tilted to nearly horizontal to look at geostationary satellites (at low elevation in northern latitudes). Used on most home satellite-TV dishes.

Used in: satellite TV receive dishes, satellite earth stations, radio telescopes (the 305 m Arecibo dish, until its 2020 collapse; the 500 m FAST dish in China; the 70 m DSN dishes), microwave relay links, cellular backhaul, deep-space communications, planetary radar (Goldstone Solar System Radar maps asteroids).

6.6 Lens antennas

A dielectric lens does for radio waves what a glass lens does for light: focuses by refraction. They are mostly a niche, used in millimeter-wave radars where the lens can be smaller than an equivalent dish (the index of refraction lets the lens focus more sharply than a parabola of the same diameter at the same wavelength). Some 60 GHz and 77 GHz automotive radars use Rotman lenses to form multiple beams simultaneously.

6.7 The microstrip patch antenna

A patch antenna is a flat metal patch on a dielectric substrate above a ground plane. The patch is roughly $\lambda/2$ on a side (in the dielectric, so smaller than $\lambda_0/2$ in air).

        feed
         │
         ▼
   ┌─────────┐
   │  patch  │    ← top metal
   ├─────────┤
   │ dielec  │    ← substrate (FR-4, RT/Duroid, ceramic)
   ├─────────┤
   │ ground  │    ← bottom metal
   └─────────┘

Properties:

• Compact (just over $\lambda/2$).
• Cheap (printed on PCB).
• Modest gain (~6 dBi for a single patch).
• Narrow bandwidth (a few percent, limited by the patch's high Q).
• Polarization linear by default; circular with two feeds 90° out of phase, or with diagonal trims of the patch.

Used in: GPS receivers (often patches), Wi-Fi access points, drone telemetry, RFID UHF tags, satellite TV LNB feeds, almost every modern wireless device. Easily arrayable, since patches can be printed side by side on a PCB and fed by microstrip transmission lines, making them the natural choice for phased arrays in 5G mm-wave base stations and automotive radar.

The Wi-Fi router on your shelf likely has 2 or 3 patch antennas inside, oriented to give complementary polarizations and pointed in different directions to help with multipath.

The PIFA (planar inverted F antenna) is a folded patch with a shorting pin, which makes the antenna physically smaller than a half-wave at the cost of a more complex feed. PIFAs are inside almost every smartphone because they fit in the bezel near the top edge while still resonating at GSM/LTE/WiFi/BT frequencies. A modern phone has 5 to 10 PIFAs and patches handling 4G, 5G, Wi-Fi, BT, GPS, NFC, and UWB simultaneously.

6.8 Slot antennas

A slot is the dual of a dipole: a slit cut in a conducting sheet. By Babinet's principle, if a dipole has impedance $Z_d$, the complementary slot has impedance $Z_s = \eta_0^2/(4Z_d)$ — about 488 Ω for a half-wave slot in an infinite ground plane, since $\eta_0 \approx 377$ Ω.

Slot antennas radiate from both sides of the sheet and are easy to integrate into metal panels (aircraft skin, car body, microwave waveguide walls). The "slotted waveguide array" used in radar is a long waveguide with periodic slots cut into it; each slot radiates, and the resulting linear array gives a high-gain fan beam.

6.9 Fractal antennas

A fractal antenna uses a self-similar geometry (Koch snowflake, Sierpinski gasket, etc.) to pack more wire length into a small area. The self-similarity gives multi-band behavior: the antenna resonates at frequencies corresponding to each scale of the fractal. So one antenna can cover, say, GSM 900, GSM 1800, and Wi-Fi 2.4 GHz with a single physical structure.

Used in: multi-band cell phones (some), tactical military radios, body-worn devices where space is at a premium. The fractal topology can give compact multi-band performance but is harder to design than a clean patch.