Wifi upgrades - ready for FTTP

[OldSkoO1]

FTTP is coming, probably 300Mbps min so I want to make sure my devices achieve that.

House is 3-storey, router/wifi-ap is on floor 2 landing, so quite central, but signal is poor on floor 1 and 3 especially at the sides.

Plus, lots of 2.4ghz noise, so 5ghz needs to be everywhere ideally.

So I have an electrician coming to route cat6 into the ceiling of all 3 floors. On one end will be a POE managed switch, on the other end will be 3x TP-Link EAP245's + a Omada cloud controller. Old router is going.

I hope I've made the right decision here. I'm currently trying to test things. This is where I need your help.

The ceiling mounted AP's are not fixed up yet, so I've just ran cables for 2 AP's now, 1 on floor 2 and another AP on floor 3. 5ghz is now excellent, noise is gone, Tx Rates are 702Mbps and up, 585Mbps lowest I've seen. Omada controller is amazing to use, band steering really works and I'm now trying to test performance with iperf3.

FTTP is going in next month. I'm trying to simulate fibre modem to wifi client by running iperf3 from an ethernet connected pc to a wifi laptop. I'm getting 550Mbps - 650Mbps. Does this seem correct? And will it accurately simulate fibre modem to wifi client peak possible performance?

If so, and I haven't misunderstood anything, this has been a worthwhile upgrade and I may be able to utilise a 500Mbps fibre connection.

P.s. I assume when I test Wifi to Wifi on iperf3 the rates are halve as the available bandwidth is split to accommodate both clients. I did however try to do this same test between AP's and despite them having a dedicated backhaul, wifi to wifi with client and server being on separate AP's still only gave 350Mbps. Not sure why.

 

[Avalon]

Before you get too carried away, consider that accepted logic is that anything that can be wired, should be wired - wifi is for IoT and mobile/streaming devices and in terms of bandwidth, most of those devices have relatively small levels of local storage, also an 8Mbit video stream is still an 8Mbit video stream regardless of it being over an 10Mbit connection or 10Gbit.

So what's the usage case that's driving the need for full speed wifi? Your testing seems reasonable to simulate LAN throughput, it doesn't take account of routing/peering on your WAN connection, the remote servers ability to send data, or - potentially more importantly - your client's ability to handle/process it. For example, I have a remote box on a symmetrical gigabit connection, Xeon 1230v2, 2x2TB in RAID0 (mechanical), I can push 100MB+ each way easily in benchmarks, but if it's got to process data (eg write data, unpack it, verify it and then repair it) while doing that, it's going to suffer, IO/CPU/RAM become the limiting factor in that order, before I can saturate the WAN. Iperf3 testing won't tell you any of that.

 

[WJA96]

Those numbers seem incredibly optimistic for a 2x2 AC access point. TP-Link claim it’s a 1300Mbps AC access point and they run half-duplex so the absolute theoretical maximum you’ll ever see is 650Mbps. That’s just a fact. And if you drop a 1x1 5GHz client on the network (5GHz n device) the best you’ll see is half that again, because all clients have to run at the speed of the slowest client. So either you’re in WLAN nirvana (which is feasible) or the software you are using is adjusting for half duplex/full duplex. In reality, if you’re seeing speeds north of 300Mpbs you’re doing well. And as Avalon has stated, get as many devices as possible off the WLAN onto wired LAN because it’s the TVs and Xbox’s that slow down and hog your wireless bandwidth.

One last thing to bear in mind is that when testing a fast connection (and 500Mbps IS a VERY fast connection, the rest of the internet often isn’t capable of feeding a 500Mbps connection. In our office in Salford Quays we have a 2Gbps leased line. And it almost never runs over 1Gbps actual throughput because it’s throttled by the incoming stream from other websites. Even people you’d expect to have VERY fast internet like YouTube, Facebook, Apple and Microsoft use QoS to stop ultra-fast connections from taking up too much bandwidth.

It does sound like you’ve done the right thing though and I suspect you’ll be very happy with your system.

 

[OldSkoO1]

Thank you both for your advice.

I agree, my obsession with maximising speed isn't a realistic requirement for typical activities. I think it has come from suffering with high levels of jitter and noise from the crowded wireless environment, the make up of the house (block dividing inner walls) and the inability for 5ghz to really cover the property.

I'd regularly see just 20-30Mbps in many places of the house, even when I don't need 20Mbps, which is perfectly suitable for UHD streaming, there's no way a UHD stream would be consistent, because of the jitter and interference. Simple wifi streams between devices in the house would get interrupted, VOIP calls would breakup/disconnect or drop. Its not good unless you are close to a good 5ghz signal. And this is with a 70Mbps WAN, soon to be 300Mbps.

So I was keen to try and get a clean 5ghz environment, well configured, good coverage and no overlaps or interference. And my yardstick was iperf3 testing in multiple locations to see the deviation from a strong signal max throughput.

Today I've played with power settings and I have
AP1-Floor1: Channel 36
AP2-Floor2: Channel: 52
AP3-Floor3: Channel 36

AP3 was creating some interference with AP1. Since turning down the power things look solid. iperf3

Ethernet-wifi: 630Mbps
Wifi-Ethernet:740Mbps with a peak of 770Mbps

Again I know I'm fixating on theoretical max speeds that in reality aren't needed when even a UHD stream doesn't need much more than 24Mbps peak, but I think it gives me some sense of achievement knowing Ive fixed the problem from years of annoyance, poor wifi stability and knowing that on the occasions where moving large files from the internet, most devices will be able to utilise the majority of a 300Mbps connection to do so.

 

[OldSkoO1]

iperf3 numbers

wifi to ethernet
Connecting to host 192.168.1.113, port 5201

[ 5] local 192.168.1.107 port 55706 connected to 192.168.1.113 port 5201

- - - - - - - - - - - - - - - - - - - - - - - - -

[ ID] Interval Transfer Bitrate

[ 5] 0.00-10.00 sec 1009 MBytes 846 Mbits/sec sender

[ 5] 0.00-10.01 sec 1008 MBytes 845 Mbits/sec receiver



ethernet to wifi
Accepted connection from 192.168.1.113, port 55596
[ 5] local 192.168.1.107 port 5201 connected to 192.168.1.113 port 55597

- - - - - - - - - - - - - - - - - - - - - - - - -

[ ID] Interval Transfer Bitrate

[ 5] 0.00-10.01 sec 767 MBytes 643 Mbits/sec receiver


wifi to wifi - same AP
Connecting to host 192.168.1.113, port 5201

[ 5] local 192.168.1.122 port 55846 connected to 192.168.1.113 port 5201

[ ID] Interval Transfer Bitrate

- - - - - - - - - - - - - - - - - - - - - - - - -

[ ID] Interval Transfer Bitrate

[ 5] 0.00-10.00 sec 496 MBytes 416 Mbits/sec sender

[ 5] 0.00-10.01 sec 495 MBytes 415 Mbits/sec receiver


wifi to wifi - different AP
Connecting to host 192.168.1.113, port 5201

[ 5] local 192.168.1.122 port 55941 connected to 192.168.1.113 port 5201


- - - - - - - - - - - - - - - - - - - - - - - - -

[ ID] Interval Transfer Bitrate

[ 5] 0.00-10.00 sec 619 MBytes 520 Mbits/sec sender

[ 5] 0.00-10.00 sec 619 MBytes 519 Mbits/sec receiver