Mesh Wifi: Does it Live Up to the Hype?
The past few months or so, in geek forums there has been discussion and chatter about mesh wifi. People ask: “is it any good?”, “what’s the difference?”, and “what problem does it solve?”.
Homes are growing rapidly with a large number of devices hungry for (reliable) wifi.
Some of these are:
– smart doorbells – smart lights
– alarm systems – cameras
– laptops – ipads
– phones – watches
– TVs – refrigerators
– ovens – washing machines
… and so forth
The array of wifi-connected equipment is large, and growing fast – each demanding their share of that connectivity.
As an engineer, I tend to look at hard technical facts, rather than glossy marketing messages. So, I spent weeks, if not months reading, considering and digesting the information available. Utilizing various sources around the net, searching for reviews, comparing comments and specifications, and trying to correlate the marketing message with actual technical specs: what are we being sold?
Finally, when half of the family took off to a hockey tournament, I agreed to spend some time on home improvement, and got some new hardware (which technically may not be the actual truth – these products today have more software than hardware, if the research and development effort is considered).
Description of the Problem
We live in a three storey house, lowest storey being an underground basement. I upgraded the house with Cisco WAP371 Dual MIMO Radio access point in 2015. The access point was strategically installed into the center of the ceiling in the upper floor. Additionally, there was a router access point and another older Cisco access point in the basement – nothing in the main floor.
This setup was serving the purpose reasonably, until the Cisco access point in the basement died. So, we lost wifi coverage in the large part of the basement, and also half of the main floor located on top of the access point. This was naturally a major degradation to the wifi coverage and performance.
The wifi performance had come to the point, where connections were not reliable, and some areas had no proper coverage at all. Something had to be done. (…to silence the family complaints.)
The project steps
First of all, the Linksys Velop AC2200 is not actually tri-band, as advertised, it’s tri-radio – it has one radio for 2.4GHz band and two radios for 5GHz band. According to specifications; Tri-Band AC2200 (867 + 867 + 400 Mbps) with MU-MIMO and 256 QAM. Each node has two ethernet ports, so it is possible to daisy chain the network equipment, and even use one or both ports to connect wired devices to the network.
Linksys Velop by default also works as a router. But in my case, I could not replace the ISP provided router. So I just set the Velop on Bridge mode, and used it as an extension of the existing network.
Linksys Velop was on sale at https://www.bestbuy.ca , so I went on to purchase Linksys Velop AC2200 Triband bundle of three nodes, and got them installed.
The results are simply amazing.
After installing the first node into the basement, connected with gigabit ethernet to the local LAN, we saw immediate downlink speed increment all the way upstairs. This is remarkable, as there is the whole main floor between the node and the device.
The Cisco WAP371 is unable to achieve the same speed, even if it’s on the same floor. This is a clear testament to the wifi radio technology advancement, during the past few years.
Running more of the tests and implementing the rest of the nodes, one node on each floor, without direct lan cable, utilizing the Velop radios for backhaul: the downlink speeds were hovering around 65Mbps. That’s already good enough for general use, but I wanted to know what’s the maximum speed Velop can actually deliver.
So, I needed to wire all of the nodes with a gigabit lan cable to push them to the limit. Routing the CAT6 in the house was the biggest job of the project. And I didn’t complete it without some extra unplanned patch up work! But, as the walls and ceilings are done by drywall, they are easy to work on.
I also realized, my internet connection is only 100Mbps. This meant I also needed an iperf server on the local network, plus an iperf client on my iPhone 8 to work around the ISP speed limit and truly measure the speed of the solution.
This was pretty straight forward, as I already have Linux server running in the network. It was as simple as using yum to install the server, start it and get going with the tests.
Finally after completing all the tests, I removed Cisco WAP371 access point, and disabled the wifi on the ISP provided router, resulting in one single wifi network/SSID visible to client devices.
The Test Results
The following table describes the downlink speed test results in Mbps for 8 different locations for 3 different wifi networks.
|Main Floor||Area 1||11||0.2||339|
|Main Floor||Area 2||16||0.2||405|
Linksys Velop clearly outperforms the other two networks, hands down. It also provides seamless coverage at every nook and corner of the house, which was simply impossible to achieve with the old/existing hardware.
Linksys Velop is a great system, which allows users to get seamless wifi roaming in the whole house, without pulling any additional wires. There is only one single wifi network, or SSID, visible and users simply attach to that wifi network. The Network serves client devices on 2.4GHz and 5GHz band, and roams seamlessly within the house, with no interruptions in the wifi.
Yes, it’s a bit of a pricey system, but there are no monthly fees, like some other alternatives. Plus it doesn’t require you to use or hook up on any additional services.
In order to get the full benefit, each node needs to be wired with gigabit LAN to the router, as the wireless backhaul will not be able to match wired gigabit connection.
Linksys Velop is not POE compatible, so I went on to purchase two sets of PoE Splitter TL-POE10R and PoE Injector TL-POE150S. The injector feeds the gigabit lan cable with data and also power, while the splitter at the node-end separates the power and data from the CAT6 cable, and I am able to run Velop nodes at the location, where only ethernet socket is available.
Total Cost of the project was around $500
– Linksys Velop being on sale
– 2 sets of POE equipment
– 30 meters of CAT6
– 8P8C connectors
And some creative work getting the CAT6 routed inside false ceilings and walls.
– Excellent Radio performance
– Ability use node as router and a lan switch
– Seamless roaming
– True whole house coverage with performance.
– No ceiling mounting option
– no built-in support for POE.
Overall, I am impressed by the Linksys AC2200 Velop performance. Though I did need to run CAT6 for each node, in order to get the full benefit of the potential speed gain available.
– TP link POE injector and splitter is not providing enough juice for Velop nodes, so I have replaced those with D-Link POE injector DPE-301GI and splitter DPE-301GS, and now the nodes are booting up properly.