I was halfway through a wireless survey when the client proudly told me they’d solved their Wi-Fi problems by installing three more access points.
They had.
For about a week.
Now users were complaining again. Video calls stuttered. Laptops randomly slowed down. Nobody could explain why the signal looked excellent but performance was worse than before.
The problem wasn’t coverage.
It was that every access point was competing with every other one.
I’ve seen this more times than I can count. People assume poor Wi-Fi means they need more access points. In reality, many networks already have too many.
More radios don’t create more air
Unlike a wired switch, Wi-Fi is a shared medium.
Every device on the same channel has to take turns transmitting. As you add more access points using the same frequencies, you’re not creating more capacity. You’re creating more devices waiting for their turn to speak.
That’s co-channel interference.
The access points aren’t broken. They’re doing exactly what they should. They’re politely waiting for the channel to become available before transmitting.
The result is a network that looks healthy on paper but feels slow to everyone using it.
One of the first things I check when I’m on site is whether neighbouring access points are talking over one another instead of helping each other.
Bigger channels aren’t always better
Channel width is another area where people chase bigger numbers.
An 80 MHz channel can absolutely deliver higher throughput than a 20 MHz channel. Under the right conditions.
The problem is that wider channels consume more spectrum. In busy environments there are only so many clean channels available before everything begins overlapping.
In a detached home, 80 MHz might work perfectly well. In an office building with dozens of neighbouring wireless networks, it often creates more interference than additional speed.
I’ve seen small businesses running 80 MHz channels simply because that was the factory default. Dropping them to 40 MHz—or even 20 MHz in crowded environments—improved performance immediately because there was less competition for airtime.
Higher peak speeds don’t matter much if the network spends half its time waiting.
“Auto” isn’t an intelligent design
Automatic channel selection has improved over the years. It’s still not something I blindly trust.
An access point can only make decisions based on what it sees at that moment. It doesn’t understand the layout of your building, how clients move through it, or what the neighbouring offices will look like after everyone arrives at work.
I’ve seen deployments where every access point independently decided the same channel was the best choice. Technically, each device made a reasonable decision. Collectively, they made a terrible one.
That’s why I still review channel assignments instead of assuming automatic equals optimal.
Strong signal isn’t the goal
People often show me a phone with a full-strength Wi-Fi icon and insist the network can’t be the problem.
Signal strength is only part of the picture.
A client can have an excellent signal while competing with dozens of other devices on the same channel. It can have a strong signal to an access point that’s already overloaded. It can also cling to an access point at the far end of the building because it hasn’t decided to roam yet.
Good Wi-Fi isn’t about having the strongest possible signal everywhere. It’s about giving devices a clean, efficient path to communicate. Those aren’t the same thing.
Sticky clients make good networks look bad
Roaming decisions are made primarily by the client device, not the access point. That’s an important distinction.
Some laptops and phones happily move to the strongest available access point. Others refuse to let go until the connection becomes almost unusable. Those are what most engineers refer to as sticky clients.
I’ve watched users walk from one end of an office to the other while remaining connected to an access point several rooms away. Their signal gradually deteriorates, throughput drops, and everyone blames the wireless infrastructure. In reality, the access point they should be using is only a few metres away.
Good wireless design encourages clients to roam naturally. Poor design often gives them too many equally attractive choices.
Turning up the power isn’t a solution
One of the quickest ways to create wireless problems is increasing transmit power everywhere.
It feels logical. If more signal is good, more power must be better. Usually the opposite happens.
The access points can now hear each other from much farther away, increasing co-channel interference. Clients also don’t transmit at the same power level as the infrastructure. The access point may hear the laptop perfectly, while the laptop struggles to send data back with the same reliability.
Wi-Fi conversations work both ways. Balanced coverage is usually far more important than maximum coverage. I’d rather deploy an access point in the right location than try to compensate for poor placement by turning the radio up to maximum.
Every building has its own personality
This is one reason I avoid rigid rules. Concrete walls behave differently than drywall. Warehouses behave differently than offices. Apartment buildings are very different from detached homes.
The goal isn’t to memorise a perfect channel plan. It’s to understand how radio frequency behaves in the environment you’re working with.
When I’m troubleshooting wireless issues, I’m rarely looking for a single faulty setting. I’m looking for patterns. Too many access points. Channels overlapping unnecessarily. Transmit power that’s far too high. Clients refusing to roam.
Individually, those problems might be manageable. Together, they create the kind of Wi-Fi people describe as “randomly slow.”
Design for airtime, not signal bars
The best wireless networks don’t usually have the strongest signal. They have the least unnecessary interference.
Before buying another access point, take a step back and look at how your existing ones are using the available spectrum. A cleaner channel plan, sensible channel widths, and thoughtful access point placement will almost always deliver better results than simply adding more hardware.
Questions or corrections? info@mb-networks.ca