So knowing the internals means we can get a reasonable understanding of just what we can expect in terms of performance. Fast Wi-Fi, hopefully incredibly fast routing and switching and generally being a fairly high performer. So, how does the router stack up? For testing I’ve specifically been using the device in ‘real world’ scenarios, rather than lab conditions. I can easily replicate Linksys’ own performance metrics that way – but let’s be honest no one will be using a router in lab style testing scenarios where everything is as close to perfect as can be.
Initially let’s take a look at uptime. I’ve been using the router as my connection to the web since it arrived, though a few config changes here and there meant the system had to be rebooted sadly so the router uptime and authed session uptime isn’t going to be as long as I’d like, still I should be able to confirm roughly when I rebooted and see if that tallies with when my ISP has seen sessions drop on their RADIUS platform.
|Start Time||End Time||Downloaded Data||Uploaded Data||Disconnect Reason|
|2016-07-17 18:08:09||2016-07-18 10:21:25||159.9GB||19.7GB||Idle-Timeout|
|2016-07-06 00:26:26||2016-07-17 18:06:49||526.6GB||221.5GB||User-Request|
|2016-06-18 22:03:00||2016-07-06 00:25:06||901.2GB||477.4GB||User-Request|
|2016-06-17 14:29:48||2016-06-18 22:01:38||11.5GB||6.4GB||User-Request|
|2016-06-17 14:17:25||2016-06-17 14:18:11||49.9KB||29.3KB||User-Request|
Unfortunately, I can only obtain sync history for the last 14 days as further data is not provided by Openreaches FttC diagnostic tool that I am aware of, however from checking the graph it shows only a single drop on the 18th, which would have been caused by a loss of power – this corresponds to the Idle-Timeout recorded on my ISPs RADIUS platform, since my session would have simply timed out after several minutes of inactivity. Thankfully, the RADIUS recording other disconnect reasons as being User-Request implies the platform has received what is known as a ‘Dying Gasp’ from the Linksys device, that is to say a message saying it’s either going down due to shut down or reboot, which would tie in to me making changes to the config that require hardware reboots. Either way, the connection has been pretty much rock solid, with no unexpected auth drops in the month I’ve been using it.
Next up, we take a look at the Wi-Fi performance of the device. Sadly, because MU-MIMO is still relatively new, I lack devices which support it natively, however as mentioned earlier, Linksys have shipped several WUSB6100M USB dongles, these currently only support Windows operating systems however, meaning I actually only have a single device I can use them in that will see the full benefit, since all the other Windows devices I have access to currently support USB2.0 only. As such, I can only really test raw throughput and not the full MU-MIMO capabilities of the device. This said, due to how the technology works we can establish a rough baseline based on testing to a single networked device, so without further ado, let’s see just how it performs over wireless.
Testing was done from my desktop PC, over Wi-Fi using the WUSB6100M USB dongle to a Linux server using iPerf. The EA9500 is at the opposite side of the next room to me, meaning that there is an interior wall and a fair bit of distance to contend with, far from optimum conditions, but then most users will never be able to replicate lab tests anyway.
First up, we test the 5GHz connection
Despite the less than ideal conditions, the wireless AC is screaming along at a decent clip, this certainly shows just what the high end internals used in the EA9500 are capable of in a normal household. In terms of MU-MIMO functionality, since there are two 5Ghz Radios, it’s fair to image that the same performance would easily be replicated over a second device connecting to the AC connection also.
Next up is the 2.4Ghz.
Next up is the 2.4Ghz.
While this is much slower than the 5Ghz connection, this is to be expected. Even so, it’s still no slouch here, providing speeds well within expectations and pushing the higher end of what I have seen most devices providing 2.4Ghz antennas in terms of performance.
Next up, we test the range and how well the network can penetrate a building. For reference I live in a three story Town house, which is approximately 10 – 20 years old. Construction is relatively cheap as is the trend with new builds in the UK, so only supporting walls are brick, the rest simply being hollow walled plasterboard with wooden frames and wood flooring. The EA9500 is located in the back corner of the living room on the second floor, again a less than ideal setup forced due to the location of my master socket. Ideally I’d normally run an ethernet cable from the VDSL Modem to where I would want the router to be positioned, but I try where possible to emulate how I think most users of the device would have things set up, so running cables to relocate the router isn’t an option, since years of front line ISP work have taught me that this is something most home users tend to be averse to doing.
As you can see, signal strength with this thing is pretty solid throughout, while 5Ghz will drop off due to internal infrastructure it seems to be holding its own here and where it doesn’t well, Linksys have a solution to the in the form of the RE7000. Similarly, the 2.4Ghz radio performs admirably too, while it may lack the range of the 5Ghz signal and does inhabit a fairly congested part of the spectrum, leading to interference to be an issue, it more makes up for it in terms of how well it can penetrate walls and other obstacles when compared to 5Ghz, similarly the RE7000 can also be used to extend the range of the 2.4Ghz signal when necessary