We put our routers through various tests. To get a general idea of the wireless throughput that a router can supply, we connect three laptops wirelessly to the router we are testing and position these in our testing environment, at 3 meter, 7 meter and 10 meter. In these tests there is line of sight to the router from the laptops. With the 802.11n tests we use Intel Centrino Ultimate-N 6300 chips, with the 802.11ac test we use devices with the Intel Wireless-AC 7260 chips. We use IXChariot, by Ixia, to run the tests. We have chosen for the High Throughput script, as that is a script that all routers can run without issues. The Ultra High Throughput script causes a collapse in throughput in many devices.
It should be noted that the 802.11ac chip that we use has two antennae connectors, so it cannot use three or sometimes even four antennae connectors that the current high end models have. It is at this moment, as far as we are aware, one of the few 802.11ac chips that can be purchased separately. If you if you want to upgrade an existing laptop, one often ends up with this product and it is used in many laptops by manufacturers. This is why we use this specific chip. As soon as we can get our hands on a MU MIMO chip that can be placed in these laptop, we will naturally upgrade them with these chips.
We still carry out our wired tests when testing routers, but we only dedicate a separate page to those results if these don't fully saturate the fully bandwidth of the gigabit connection.
We have experimented quite a bit with developing new testing procedures to map the throughput of routers in a better way. In the end we settled for a few additional tests that we carry out in the office, where we have a single BTO laptop and use it to measure the throughput speed on two locations in the building when we transfer a file from a Asustor As-604T NAS in the network via the router to the laptop wirelessly and vice versa. In the NAS we have placed SSD's which allows for wired transfer speeds of well over 100MB/s, which means that the gigabit connection will be fully saturated. We use TeraCopy to copy the file. The file is a video fragment of a little over 307 MB.
We initially carry out this test on two different locations, but we have decided to carry out the test in five different locations for this test. The left-over two give a good idea of what can expect in practice. Copying a file won't always be the smoothest process, due to things like processes running the background on the laptop. Because of this, we run the tests multiple times on different points in time, after which we will omit any outliers, if they are found.
The first test is done at 10 meter with line of sight, in order to determine the throughput in nearly optimal circumstances. This result is always a little lower than what we measure with IxChariot, because of the increased overhead. For the second test we place the laptop in a room above the router. It is directly above the router, but there’s a thick concrete plate in between them. The distance between the router and the laptop is about 4 metres in this test. That might not seem like much, but in practice it has proven tricky enough for the routers we have tested so far.