It is so confusing trying to measure your ISP Internet Router/Gateway WiFi speed around your house.
Recently moved from Optus (atrocious customer support) to AussieBB (customer support superior & Friendly). I also moved my NBN from FTTC to FTTP and went to NBN UltraFast 1000/50… With a Netcomm NF20MESH Gateway & NS02 Mesh Sattletie.
Still trying to measure my WiFi speed around the home is a nightmare…
Using my iPhone mobile in different spots around my two storey home I get different very fluctuating upload/download speeds at the same spots (close to router & at other spots) at different times. The home WiFi speed is different to the WiFi speed of the Internet at the Gateway/Router. Also - if you use the common Ookla - SPEEDTEST and the AussieBB Speed Test - at the same time - you get big differences.
Understand speed tests check speed between your WiFi Address (your router) and some other external ISP location to test speeds… So this may explain the variable fluctuating AND CONFUSING WiFi speeds around the home.
Surely - measuring WiFi as accurately as possible at different DEFINED points around the house - and being able capture them in a log over time is important for anyone struggling with WiFi instability and variability.
Surely there must be some answers… So if you have solved this problem - here is your chance to help others. Also CHOICE should seriously consider this vexing consumer challenge with a vital & key utility for many people that rely on stabe/fast internet for their livelihood…
Channels being used particularly in the 2.4 GHz range can have effects on speeds seen throughout the day, from your post it would seem you are quite familiar with IT and this may be something you are already quite au fait with. 5G because of the number of channels shouldn’t be seeing large differences. As you are using a Mesh, you hopefully have placed the devices in the best postions to get best coverage. Though if two story you may need a further Mesh unit or two or even perhaps a powerline adapter. I have in my single story house a 3 unit Mesh network, I would imagine a 2 unit Mesh would be possibly not adequate enough, particularly with walls, wiring, metal, and so adding to the blocking of signals.
To create a heat map if you haven’t done so already will be a big help in getting some clarity. What is Wi-Fi Heat Map? - IT Glossary | SolarWinds
For an iPhone try NetSpot WiFi Analyzer App - App Store. I use a Android phone with a different analyser that suits my needs when I need to do heat mapping. Though I usually am an iPhone user myself.
Speedtesting will vary on what server you are hitting, ABB uses it’s own servers to test against. Ookla will pick any server that seems to be best choice and may be closer than any ABB choice so giving a different outcome. If I test with the Whitebox from the ACCC (SamKnows One), I get results that are usually much different to the ABB or Ookla tests for DL speed (much better results in fact).
Ookla (to best choice)
ABB (to closest server)
SamKnows One
SamKnows RealSpeed test (tests to Whitebox and to my mobile device that requested the test)
RealSpeed (Whitebox and Desktop device)
Both the RealSpeed tests were conducted at exactly the same spot, yet the difference in device DL speeds was quite a bit different.
Yes, exactly.
You can’t speedtest via WiFi because you are measuring the combination of two things … the speed of your internet connection and the speed of your WiFi.
An actual speedtest of your internet connection should always be done wired i.e. via ethernet.
For WiFi speed testing (or indeed any speed testing around the house) I use iperf3 (which is specifically designed for testing within the local network). I only mention this because the previous post suggests that you are familiar with IT.
I would also suggest that if speed is really important to you, and given that you are paying for premium internet, it may be better to wire up (assuming that you are an owner occupier). At the very least, extending the network upstairs via ethernet, so as to have a separate Wireless Access Point upstairs could be useful. An alternative to ethernet would be ethernet-over-power-wiring but that may not be able to do justice to a 1 Gbit/s internet connection.
WiFi speed will vary depending on the construction material of your house, the area and shape of the house, and on many other factors e.g. neighbour’s WiFi activity if they are close enough.
For any local network speed testing, you need to contend with concurrent activity i.e. you need to quiesce all other devices, which in today’s world could be quite the exercise.
Also, on the occasions that I am interested in WiFi around the house, I would be inclined to look at signal strength (measured in dB) rather than speed.
Actually, this is a large and complex subject area.
Then, in my opinion, it’s a no-brainer that you would wire up.
Just as an example the TP LInk AV2000 offers up to 2,000 Mbps. The Asus 1200Mbps AV2 1200 Wi-Fi Powerline Adapter offers 1,200 Mbps. If not wired in they use some of the existing WiFi bandwidth as the carrier of the signals for the mesh system. They never have in my experience achieved maximum throughputs, but even a 500 - 900 Mbps throughput is very fast. Wired is best but many people live in buildings that to retrofit wiring would either be impossible or very expensive, hence powerline adapters or WiFi meshing as next best options. Powerline adapters need to be on the same circuit as well, this may further limit the possibility of their use.
I very much agree with
From what I can find online the Mesh20 units achieve aound 400 - 600 Mbps throughput out of the room where it is located. Building construction will affect performance. Why getting a heat map done would be very helpful in determinig positioning and number of mesh units to get best possible performance. Another setting which might be slowing individual throughput is having OBSS Coexistence turned on, I am not recommending turning it off as this may affect older 2.4 GHz devices and Bluetooth devices. If using mostly 5GHz bands then turning off OBSS may improve for some devices the throughput seen.
Another possible limiting of WiFi speed is that this unit only has internal antenna, this can impact performance depending on where the signal is trying to be received at. External antennas can be positioned to alter the signal footprint, the only option when boxed is to rotate the box to see if the signal is improved, and this can mean signal is reduced in other ares where it was stronger before. I prefer multi external antenna units over in the case ones.
The other is that with most mesh systems everything is transmitted twice, they have a speed about half the speed of that if the WiFi connection was direct to the router.
I like @person point about etherneting the house and installing WAP. This retains router type WiFi speeds.
If you live in a heritage-listed house 
then, yes, there may be serious impediments that can’t be overcome.
If you live in a strata-titled property then, again, there may be significant impediments.
Let me rephrase then: after eliminating the impossible situations if you can, in my opinion, it’s a no-brainer that you would get a qualified data cabler in to discuss your requirements and give you a quote for the job.
Yes, it may be that ultimately the quote is higher than you are prepared to pay and then you would look at cheaper but less performant options.
I do have one powerline adapter in use, and it functions well. However the actual measured performance (iperf3) is only a fraction of the marketing speed. (That isn’t really a problem for me because the internet connection isn’t exactly setting the bar high.) The performance - for a given AV standard - could depend on a heap of factors e.g. length of house wiring involved, age of wiring, arrangement of circuits, number of powerline units.
With either powerline or WiFi you are sharing bandwidth within the house, which is an additional challenge with such a fast internet connection.
Hi All
Some great feedback - so thank you for experience and advice to
@grahroll plus others
I have a bit of work to do to:
- Look at apps to get WiFi “heat map” for my home
- Look at further Ethernet hard wiring - a bit difficult/expensive in my existing older two storey house
- Look at better mesh routers that the ABB Netcomm - with external antennas & additional mesh satellites - to position based on heat map
Again thanks guys for prompt responses
Will post update on outcomes of my TODO list
Never say Never, but just how fast an NBN or internet service do most require in the typical home. There will always be exceptions to the norm. For those who have a genuine need for the fastest fibre connections, they do not come at a low cost. The interests or personal business needs that might require Gigabait connectivity likely require a substantial investment in the devices/digital equipment in use. Perhaps not such a stretch to follow the previous mentioned,
Further to the comments that Mesh networks share bandwidth. Hence any in home networks dependant speed test results can be influenced by other activity or nearby interference. Reviews of Mesh products by popular online publications also mention that some products are better at how they optimise the connection to a device that’s is constantly moving around the premises. The theory that the nearest Mesh node with the best connection (IE fastest) will always be the one the Mesh network routes traffic through is not a given.
One useful resource for general and individual product reviews.
We have a lot of data usage in our home. Nothing for us to see between 4 to 7 TB a month of downloads and around 2 TB up. Nothing like many businesses see but certainly a lot for most normal households I would think. Decent speed becomes almost a necessity for reasonable perfomance at those levels.
When the 2 Gbps plans are somewhat more common we are anticipating a move to that, our house was designed around an eventual 10 Gbps connection. Initial NBN Co and retailer discussions concluded in June this year. Hopefully a rollout within this financial year.
For sure - but the OP made three specific observations
- the OP has a 1 Gbit/sec download plan
- the OP in some sense earns a livelihood at home
- the OP gets inconsistent results attempting to measure speed via WiFi between a specific point in the house and the internet, for various points
and I chose to accept the need for that speed as a given - and the advice I gave in response, considered in its entirety, would be largely the same with a substantially lower speed internet connection e.g. 100 Mbit/sec download.
And, yes, I agree. If I were bothering to pay for a 1 Gbit/sec download plan, it needs to be matched with decent equipment, thoughtfully chosen and arranged and configured.
For the OP suitable responses provided. Noted also the reference to potential livlihood needs.
The topic heading is non specific. It will also attract the less demanding user. What might more typical households look like?
More typical household may also find benefit in the advice for Mesh products. There are varying price points and speeds. Lower cost products include range extenders and power line adapters which may also be adequate for most households. They are a straightforward add on to existing hardware. All can benefit from optimising the location of the network WiFi router/access points.
We’ve used both and found a single range extender or powerline adaptor adequate. Albeit 200GB per month and internet sub 100 Mbps has been our norm. Prior a forced diet of speeds of 2-12Mbps until relatively late in the NBN roll out. Apologies, it may explain in part and difficulty in relating to Gbps internet speeds.
Prior to the NBN we used a router with a range extender (WiFi + 4x Ethernet ports) located in the home office. Experience is the iPhone (14) is not the best device at connecting to the home WiFi. The OP mentioned using an iPhone when testing. The iPad, PC and older iPhone 8+ all out perform the iPhone 14 when connecting to the home WiFi. If it was necessary to produce a heat map of our home WiFi which device might one prefer?
Some more great advice & discussions - thanks Guys
By the way in the USA Google or fiber.google.com (GFiber) offers in some USA cities the following…
I guess it is only a question of time…
A decent Android phone 
Apple lock up a bit of data that would make the heat mapping software a lot more useful. Apple doesn’t allow developers (of the software) access to the WiFi data which cripples usage. To get best use from an Apple requires jailbreaking which definitely is not something I recommend or do.
A heat map will produce a nicely coloured display of signal strength throughout the area that is surveyed. Any areas of less satisfactory coverage will show whether an iPhone or an Android one is used. Difference is that the Android apps provide a much better heat map and data points than any Apple product ever can.
To produce the best results most apps work best if the user has uploaded a floor map of the area (with measurements of dimensions) then the user walks around tapping the app display to take the readings (most require short taps, some require long presses). More data points makes for a much more complete mapping. Some apps even suggest the best places to put APs (access points) to improve coverage when either mapping or after the map is finished.
Mapping should be done where possible with only the main router running and turning of any satellites for the first round. Then turn on satellites after placing them where a user considers the best locations (or where recommended). Remap to see if the placement has given the result wanted/required. In the situation posed by @Anthony007 with only one satellite, placing where it is recommended or desired will show if another satellite is needed or desirable. If another is needed, then place both satellites where required and remap. My estimation is that at least one more or even two satellites will be needed for best results. It is possible that even 3 or 4 more may be required, it is just a guess until some mapping has been done, noting we don’t know what size the home is, what materials it is made of, what is in the interior e.g. furniture, metals, water (such as aquariums), wiring, number of walls and other things that may impede signals.
If using free versions of Apps then some limit the number of projects. A two story house would mostly be two projects (ground and first floor). This for most users will be suitable even if just using the free Apps.
@mark_m, considering even our heavy usage here WiFi meshing is more than adequate, we tried both WiFi with backhaul and using hardwired APs. Hardwired is the best outcome but even at pure WiFi we had 400-500 Mbps coverage throughout the house. While we have a small yard, before other buildings were built around us our coverage with very strong signal strength extended quite some distance from our premises. This was around 17 metres from any external wall for almost perfect strengths. For one neighbour when they first moved in, before they got nbn connected, we supplied their house with our Guest network and they had no problem with coverage.
If we had a much larger yard (acreage) we would have placed APs externally to cover the areas required. Ubiquiti have some of the best devices for coverage in my opinion for price but others such as the ASUS mesh setup we use (and definitely not limited to this brand) give very good results for households.
Considering that many household have 4 wired ports on their routers of 100 Mbps, but often have much faster WiFi connections (even taking into account the superior performance of wired), then even with purely WiFi meshing they won’t suffer. If they upgrade to 1 Gbps ports (even then they are mostly 4 port) then wired would make best sense to a device or AP, again though WiFi speeds of 400 Mbps or thereabouts is very good speeds for almost every usage even business, except where the intent is to supply coverage to many guests/workers. If the intent is to supply coverage to many more people than a normal household, then getting into business plan speeds and service requires a more expensive and professional approach.
An important point to remember is not to put a WiFi meshed AP into a poor signal area, the AP should be placed in a strong signal area to ensure best communications, it will then extend that strong coverage to poorer areas of signal within it’s footprint. If hardwired in, the AP can be placed in a poor signal area to improve coverage in that area.
Some recommendations that are free for Android
For a laptop
There used to be more free options available, most have now become paid versions. I have a commercial one, which provides a much better quality outcome but the above options should provide a more than adequate outcome for most home users to improve their network.
