The Practical Antenna Selection Guide: Band, Gain, Pattern, Connector

The four parameters that actually matter

Antenna datasheets contain dozens of specifications — frequency range, gain, VSWR, beamwidth, polarisation, efficiency, group delay, phase centre, IP rating, and more. In practice, four parameters determine whether an antenna will work in your system: frequency, gain, pattern, and impedance. Everything else is either derivable from these four or irrelevant to your application.

Understanding these four parameters — and how they interact with the rest of your radio system — is the difference between an antenna that performs as expected and one that becomes the bottleneck in an otherwise well-designed deployment.

Frequency: not just “covers the band”

The obvious requirement is that the antenna’s operating frequency range includes your system’s transmit and receive frequencies. But “covers the band” is not enough. You need to understand how the antenna performs across the entire range, not just at the centre frequency.

VSWR across the band tells you how well the antenna is matched at each frequency. A VSWR of 2:1 at the band edges means 11% of the power is reflected — acceptable for many applications, but problematic if your system operates at the edge of the antenna’s specified range. Always check the VSWR plot, not just the “typical” value at centre frequency.

Gain variation across the band is equally important. An antenna specified as “5 dBi ± 2 dB” across 698-2700 MHz will have significantly different performance at 700 MHz versus 2600 MHz. If your system operates at a specific frequency (say, 2400 MHz for Wi-Fi), check the gain at that exact frequency — not the average across the band.

Multi-band operation requires the antenna to perform well at multiple, widely separated frequencies. A “dual-band 2.4/5 GHz Wi-Fi antenna” must maintain good VSWR and reasonable gain at both frequencies. The design challenge is that a single radiating element cannot efficiently resonate at two frequencies with a 2:1 ratio — most dual-band antennas use separate elements or a deliberately broadband design with reduced peak efficiency.

Gain: how much does it matter, really?

Antenna gain is the ratio of power density in the antenna’s peak direction to that of an isotropic radiator. Higher gain means more range — but also a narrower beam, which means more precise pointing requirements. The question is not “more gain is better” but “how much gain do you need for your link budget?”

Link budget calculation is straightforward: Transmit power (dBm) + transmit antenna gain (dBi) + receive antenna gain (dBi) – path loss (dB) – cable losses (dB) = received power (dBm). The received power must exceed the receiver’s sensitivity (typically -90 to -100 dBm for Wi-Fi, -110 to -120 dBm for LTE) with some margin for fading.

When high gain helps. Point-to-point links (bridge between buildings), long-range outdoor deployments (km-scale backhaul), and environments with high interference (where directional gain rejects off-axis noise) all benefit from high-gain antennas. A 15 dBi panel antenna can extend a Wi-Fi link from 100 metres to 1 kilometre — if the antenna is pointed correctly.

When high gain hurts. Indoor environments with rich multipath, mobile devices that cannot maintain pointing accuracy, and dense deployments where narrow beams miss users — these all perform better with moderate-gain omnidirectional antennas. A 5 dBi ceiling-mount omnidirectional antenna provides better coverage in an office than a 12 dBi directional panel, because the signal reaches users in all directions, not just one.

Pattern: the shape of coverage

The radiation pattern describes how the antenna’s gain varies with direction. It is typically specified as horizontal (azimuth) and vertical (elevation) plane cuts. Understanding the pattern is essential for predicting real-world coverage.

Omnidirectional (dipole-like) patterns radiate equally in all horizontal directions, with a donut-shaped vertical pattern. These are ideal for base stations and access points that must serve users in all directions. The gain is moderate (0-5 dBi), but the coverage is uniform.

Directional (panel, Yagi, dish) patterns concentrate energy in a specific direction, trading coverage width for range. Beamwidth can range from 180° (sector antenna) to less than 5° (parabolic dish). These are used for point-to-point links, sectorised cell sites, and long-range backhaul.

Shaped-beam (base station) patterns are a compromise: broad enough to cover a sector (typically 65-120° azimuth) but shaped in elevation to focus energy toward the ground and reduce interference with adjacent cells. Cellular base station antennas use electrical or mechanical tilt to adjust the elevation pattern for different cell sizes.

The null problem. Every antenna has directions where gain drops significantly — typically directly above or below the radiating element. A vertically mounted dipole has a null directly above it. A ceiling-mounted patch antenna has a null directly below it. Understanding where the nulls are helps you avoid placing users or access points in those dead zones.

Impedance and VSWR: the matching problem

Every antenna has a complex impedance — a combination of resistance and reactance that varies with frequency. The antenna is designed to present an impedance as close as possible to the feed line’s characteristic impedance (typically 50 ohms). When the impedances match, maximum power is transferred. When they do not, some power is reflected back toward the transmitter.

VSWR (voltage standing wave ratio) is the standard measure of impedance match. A VSWR of 1:1 means perfect match (no reflection). A VSWR of 2:1 means 11% reflection. A VSWR of 3:1 means 25% reflection. Most antennas specify VSWR ≤ 2:1 across their operating band — anything worse indicates a mismatch that reduces effective radiated power.

Cable matters. The VSWR is measured at the antenna’s feed point. Once you add a cable, the VSWR at the transmitter end changes due to the cable’s electrical length and loss. A poorly matched antenna with a long, lossy cable may actually show a better VSWR at the transmitter — because the cable loss masks the reflection. This is not an improvement; it is wasted power dissipated as heat in the cable.

Always measure the system, not just the antenna. In production, measure the VSWR at the transmitter end of the cable, with the antenna installed in its final location. This captures the effects of cable length, connectors, and nearby objects that can detune the antenna. A handheld VNA (vector network analyser) makes this measurement straightforward in the field.

The practical selection process

1. Define your frequency — not just the band, but the exact operating frequency or frequencies
2. Calculate your link budget — determine the gain you need to close the link with margin
3. Choose your pattern — omnidirectional for area coverage, directional for point-to-point
4. Check the VSWR — at your specific frequency, not just the “typical” value
5. Verify the connector — matches your radio, or a low-loss pigtail is available
6. Consider the environment — IP rating, temperature range, wind load, mounting hardware
7. Test in situ — bench measurements do not capture real-world installation effects

Antenna selection is not about finding the “best” antenna. It is about finding the antenna whose four key parameters — frequency, gain, pattern, and impedance — match your system’s requirements. Get those four right, and the installation will work as designed. Get any one wrong, and the entire system underperforms — no amount of transmitter power or receiver sensitivity can compensate for a poorly matched antenna.