I’ve specified enough outdoor WiFi deployments at this point that I can now identify, in retrospect, exactly which decisions came from understanding the technology and which came from assuming that what works indoors scales to outdoors with only the antenna swapped out. The latter assumption costs money and time in almost every case where I’ve made it.
What follows is what I actually needed to know before those projects, organized around the questions I should have been asking rather than the ones I was asking.
IP rating is the floor, not the finish line
The first thing people look at on an outdoor antenna spec sheet is the IP rating. IP67 means dust-tight and waterproof to one meter. IP66 means dust-tight and protected against heavy water jets. Both are adequate for most outdoor installations. But treating IP rating as the end of the environmental discussion is a mistake.
What IP ratings don’t tell you: UV resistance of the radome material, operating temperature range, resistance to salt spray if you’re anywhere near a coast, and long-term seal integrity through repeated thermal cycling. A housing rated IP67 on day one may develop micro-fractures in the seal after two years of summer heat and winter cold. In my experience, the antennas that develop problems at year three or four in harsh environments are often ones that passed IP testing without issue at installation.
The questions I now ask before specifying any outdoor antenna for a permanent installation: what is the operating temperature range, what is the UV rating on the radome material, and does the manufacturer have data on seal integrity after thermal cycling testing. Manufacturers who can answer all three have usually put more engineering into the product than manufacturers who can only point to the IP number.
Gain figures need context to be useful
An outdoor omnidirectional antenna with 8 dBi gain sounds better than one with 5 dBi gain. In some applications that’s true. In others, the higher gain antenna creates a problem.
Omnidirectional antennas achieve gain by compressing the radiation pattern vertically. A theoretical isotropic antenna radiates equally in all directions — a sphere. A real omnidirectional antenna concentrates that energy into a flattened disk, trading vertical coverage for horizontal range. The higher the gain, the flatter the disk. A 2 dBi antenna has a relatively round pattern. A 9 dBi antenna has a very flat, narrow vertical beam.
The consequence: if you mount a high-gain omnidirectional antenna on a rooftop and need it to serve devices on the ground directly below, the vertical beam may not reach them well — or at all in the nulls. I’ve seen 9 dBi antennas installed on a 6-meter mast for a parking lot deployment where the first 15 meters of ground around the mast had noticeably weaker signal than points 50 meters away. The antenna was performing exactly as specified. It just wasn’t the right specification for that geometry.
As a rough guide: for close-range coverage where devices may be near or below the antenna mount point, lower gain (2-5 dBi) preserves a wider vertical pattern. For long-range coverage across flat ground where you need distance rather than overhead coverage, higher gain (6-9 dBi) works better. That decision needs to happen before you order the hardware, not after installation.
Polarization is almost never discussed and matters more than most people think
Most WiFi client devices — phones, laptops, IoT sensors — use vertically polarized antennas internally, though some diversity antenna designs cover both polarizations. Outdoor infrastructure antennas are usually vertically polarized as well, which works fine in free-space environments. But in environments with significant multipath — buildings, vehicles, structures that reflect RF — cross-polarization losses and polarization rotation effects on reflected signals can introduce unpredictable dead spots.
For general outdoor deployments, vertical polarization and a co-polarized outdoor omnidirectional WiFi antenna is the correct starting point. The case for a dual-polarization or slant-45 polarized antenna arises in high-multipath environments with many client devices at varying orientations — a crowded outdoor event venue, for example, or a deployment where mobile equipment is frequently tilted or mounted at non-standard angles.
This isn’t something most project specs even mention. It becomes relevant after you’ve done a few deployments where coverage maps didn’t match predictions and the root cause turned out to be polarization mismatch in a high-reflection environment.
Cable loss is where most outdoor antenna budgets go wrong
An outdoor antenna typically lives some distance from the access point or radio it’s connected to. That cable run has loss — roughly 0.1 dB per meter for LMR-400, more for thinner cables. A 20-meter cable run on LMR-240 can easily add 4-6 dB of loss between the radio and the antenna. If your link budget assumed the antenna gain would extend coverage by a meaningful margin, half of that gain may have already been consumed by the cable before the signal reaches the antenna.
The solutions are either shorter cable runs, lower-loss cable, or moving the radio closer to the antenna. In practice, placing a PoE access point or small radio unit at the antenna location and running Ethernet (which has essentially no signal loss at WiFi frequencies) rather than coax is often the right answer for longer runs. The antenna becomes the last 50 centimeters, not the far end of a 30-meter cable run.
I’ve seen installations where someone paid a significant premium for a high-performance antenna and then undermined it with the cheapest available coax in a 25-meter run. The link budget math doesn’t forgive that kind of inconsistency.
Mounting and grounding are not afterthoughts
An outdoor antenna mounted on a non-conductive surface with an improper ground plane behaves differently from the same antenna mounted on a grounded metal structure. Many omnidirectional antennas are designed with the assumption of a proper RF ground plane behind the radiating element, and without it, the radiation pattern can shift — including downward, which is rarely what you want.
Lightning protection is a separate but related issue. An outdoor antenna is essentially a conductor on a stick, and in environments with lightning risk, an unprotected antenna system can conduct a surge directly into the access point or switch it’s connected to. A proper installation includes a grounded surge arrestor at the building entry point for any cable that runs from an outdoor antenna to indoor equipment. This is specified almost nowhere in the procurement discussion and omitted in a meaningful percentage of installations I’ve seen, until something gets fried.
None of this is proprietary knowledge. It’s in the installation guides. The gap is that procurement specs rarely include it, so the people making the buying decision never see it, and the people doing the installation may or may not check.
What I now put in every outdoor WiFi antenna spec
Operating temperature range and UV rating alongside the IP rating. Gain with explicit confirmation that the vertical beamwidth suits the installation geometry. Cable type and run length with explicit loss budget. Mounting surface type and grounding plan. Surge arrestor included in scope. Those five things catch the majority of the problems I’ve seen in outdoor deployments. The antenna itself is usually the least interesting part of the decision.
