Again with the SNR
In my previous post (Eliminate Channel Dysfunction – Your Wi-Fi Will Thank You), I spoke briefly about the signal-to-noise ratio (SNR). SNR is what’s left over of the received signal (RSSI), after you subtract the ambient noise (or noise floor). For example, if the RSSI at the client device is -60 dBm, and the noise floor is at -95 dBm, then the SNR is (-60) – (-95) = 35 dB.
Many enterprise controllers show the client’s SNR as the reported signal level, while others will show the RSSI.
Too Much of a Good Thing
Knowing now how SNR works, you might be thinking, “Well, if I need better SNR, then I’ll just crank up the AP’s transmit (Tx) power.” While that does make mathematical sense, it’s not always the right move when it comes to optimizing your Wi-Fi network’s performance. Let’s get into the reasons why.
Wi-Fi Devices are Not Created Equal
In the earlier days of Wi-Fi, the primary devices connecting to the Wi-Fi were PCs, mostly laptops and some desktops using an adapter. These devices had pretty decent antennas and transmitting capabilities. Thus, it was common for a network engineer to crank the AP’s Tx power to the max, and therefore get decent coverage with fewer APs.
However, nowadays, things have changed. Devices like smartphones, VOIP (or VoWiFi) handsets, and IOT devices are commonly using Wi-Fi, and these types of devices have limited capabilities. Their antennas are much smaller, and due to battery efficiency, they are more limited in their transmit power.
Take, for example, the iPhone 5. The table below shows it’s max power capability, which changes based on the channel in use.
[ a few years ago, Jerome Henry (principal engineer at Cisco) gave a conference presentation about this same topic, and that’s where I got the iPhone 5 power data above ]
Now, considering that the max Tx power of many of today’s APs is between 23 – 30 dBm, you see the significant difference between what the AP can do vs. today’s smaller battery-operated Wi-Fi devices (roughly half).
Can You Hear Me Now?
Because client devices are more limited in their transmitting capabilities than APs, it’s not uncommon for these devices to have an asymmetrical connection, meaning if the AP’s Tx power level is higher than the client’s, then the client will “hear” the AP, but the AP may not “hear” the client. Again, this is due to the mismatch in Tx power levels, and mismatched power results in mismatched range.
Having mismatched power (and therefore an asymmetrical connection) should be avoided for several reasons. Primarily, it will result in a lot of retries.
There are several reasons why a Wi-Fi transmission (or frame) must be resent, and we’ll discuss another one here in a moment. What triggers these retries is the absence of an acknowledgement (ACK) frame from the receiving device. You see, a lot can go wrong with Wi-Fi, so the powers-that-be built into the protocol this acknowledging behavior where the receiver will acknowledge the successful receipt of a transmitted frame, by sending its own ACK frame. The scary thing is that these retries are resent over-and-over (and usually slower-and-slower) when the expected ACK isn’t received. What do you think that does to the available airtime?Eliminate Channel Dysfunction – Your WiFi Will Thank You
So, when the client’s data doesn’t reach the AP, the client resends it over-and-over again. And, the data either never gets there (resulting in packet loss), or gets there late as it tries lower data rates that are easier to demodulate at lower signals (resulting in latency).
This behavior is particularly detrimental to voice applications.
Let’s look at another conference call example (see the “Understanding SSIDs (aka WLANs) and Using Them Wisely” post for the first one). If there’s one common theme about conference calls, it’s that there’s usually at least one loud-mouth, and one whisperer. The loud-mouth is at a near scream (like an AP at 23 dBm) causing folks to pull away from their speaker or headset. However, the whisperer speaks so softly (like a VoWiFi device at 11 dBm), that the rest of us can’t distinguish what he’s saying from the background noise. What should be done to resolve this? Obviously the loud-mouth needs to lower his speaking volume, and the whisperer needs to raise his, so they both can be heard at appropriate levels.
Bringing Balance to the WiFi
I think it’s obvious by now that the way to fix the mismatched power between AP and client, is to set the AP’s power level to match that of the least-capable client device that will be connecting to your Wi-Fi. Now, this may not be as simple as it sounds, as any experienced Wi-Fi engineer will tell you, because you have to consider multiple device types, multiple application requirements, and not to mention the financial budget. So, this balancing process usually takes some tuning and prioritizing to get it right.
Some vendors may provide guidelines, or starting points. Take a look at Meraki’s RF profile templates for example, and their min/max Tx power levels.
So be a good WiFi Jedi (ohh, new t-shirt idea I think), and bring balance to your WiFi.