GoKawiilRouter manufacturers make big claims about top speeds, many of which can be misleading or at least confusing when you're shopping for a new one. That's why we put every router we review through our own, independent speed tests in our test lab, where we can do more to control for variables in the environment.
CNET uses custom-made scripts to carry out our testing and data analysis. We focus on three main measurements when testing all Wi-Fi routers, including mesh systems: speed, reliability and distance. We want to know how well a router will translate the internet coming in from your provider to a wireless connection, and we want to know how far it can do it -- even with obstacles like walls and floors in the way. To capture all three of these, we conduct two types of tests: data throughput and signal strength measurements.
CNET’s testing is unique because we remove the internet provider from the equation entirely. Even the most reliable internet connections are fairly erratic: factors like time of day, network congestion and even the condition of your neighborhood’s broadband infrastructure can all affect your internet speed. To make sure these variables didn’t skew our analysis of each router, we set up a separate server to measure the raw throughput of data from the router to a device, bypassing the need for an internet connection entirely.
Learn more: How We Test Wi-Fi Routers
Data throughput tests
Here I am running a throughput test at close distance from the router. Adam Breeden/CNET
To see how fast a router will perform in real-world conditions, we measure its throughput, or the volume of data that a router actually transmits during a period of time. All Wi-Fi signals travel through 2.5GHz and 5GHz radio frequencies -- that’s why most routers you’ll see are referred to as “dual-band.” (Tri-band routers use an extra set of channels in the 6GHz band.)
To start, we conduct a wired measurement on the 2.5GHz, 5GHz and, when available, 6GHz bands to establish a baseline. This is as fast as throughput can get for the router over each band, so it’s a good point of comparison for our wireless tests. From there, we take five different wireless throughput measurements, 60 seconds each, mimicking a household with five devices connected to the router with different throughput needs. These wireless measurements are taken from three different distances from the router: 7 feet, 25 feet and 50 feet.
As any online gamer or remote worker can tell you, speed isn’t the only factor in evaluating routers. Just as important is jitter, which measures the variations in latency, or the time it takes for data to arrive at its destination. When you’re in a video meeting and there’s a delay in audio or someone’s video freezes, jitter is the culprit. We felt jitter was a better measure than latency when evaluating a router, as average latency can show up in testing as low, even when an internet connection is glitchy.
These tests also captured packet loss, which tells us whether any data failed to reach its destination during the test. You’ll often see packet loss as a frozen video screen or a garbled voice during a video call. With both packet loss and jitter, a lower score is better.
... continue reading