A
s Wi-Fi has steadily improved, and
especially with the advent of mesh
Wi-Fi systems such as the Netgear
Orbi or Google Wi-Fi, it’s easy to think that all
other forms of home networking are
pointless relics of a bygone era. It’s a
wireless world now, so why would you go
back? However, in reality, the situation isn’t
that simple.
For a start, mesh Wi-Fi systems can get
expensive pretty quickly, with each node
generally setting you back at least £100.
If you want to stretch your network all the
way to a new home office at the bottom of
a garden, you need at least three units to
get a reliable, speedy connection, even in a
modest two or three-bedroom house.
All forms of Wi-Fi are also prone to
interference that can either slow down or
interrupt your connection; if you’re relying on
it for crucial tasks, such as online gaming, this
can be a matter of virtual life or death. This
tends to be less of a concern up close, but as
you move towards the edge of the range of
any given Wi-Fi access point, or start daisy-
chaining together multiple mesh nodes,
there can be a significant reduction in the
responsiveness of the whole system.
Obvious alternatives have been around
for a while, in the form of running a big long
Ethernet connection to your PC or using
powerline technology that can run a network
connection over your home’s existing mains
power cables. But how do these systems
now compare with the latest mesh networks
or standalone routers? That’s what we’re
here to find out.
The test setup
To really dig down into just what sort of
performance and ease of use the latest
home networking systems offer, we set up
a simple test scenario. In a three-bedroom,
two-storey brick-built house with a 20m
long garden, we decided to see how each
technology we used performed at various
points throughout the home and out into the
garden. There isn’t quite a home office at the
bottom of the garden yet, but a few sunny
days allowed for some al fresco testing.
As with our router Labs (see p50 ), we
chose three key test locations for our network
systems. These consisted of a sofa 2m from
the home’s router in the front, ground-floor
room of the house, a rear first-floor bedroom
approximately 5m diagonally back from the
router, and a spot halfway down the garden,
about 15m back from the router.
The host device of each option was
plugged into a router next to the Internet
connection in the front downstairs room of
the house, with a PC attached via a Gigabit
Ethernet connection to the router. A laptop
was then either connected with a USB/
Ethernet dongle to any wired network
setups, or its internal AX1500 Wi-Fi card
was used to connect to any Wi-Fi signal.
We then tested for connection speed using
LAN Speed Test and for ping response with
netmeter.co.uk/ping-test
The former works by running a server app
on the PC, which exchanges data with a client
app running on the laptop. You can tailor the
test to whatever file size and number of test
runs you like. For each run of the test, we set
the app to send three packets of data varying
between 5MB and 50MB, depending on
the speed of the test – there was no point
in waiting five minutes for a 50MB run to
complete on a slow connection when the
result is the same as waiting 30 seconds
using 5MB files.
The ping test simply pings a reliably
quick nearby server, which in our case was
registered as being in Manchester – you
can also manually pick the server, but we
found the Manchester option reliably fast.
While this also tested the ping of the Internet
connection, which we weren’t explicitly
testing, there was sufficient variance in the
ping due to the local network connection
being used that we still obtained clear results.Ethernet
With our test setup established, it was time to
start putting our home networking systems
to the test, starting with the granddaddy of
them all (at least as far as common modern
options goes) – Ethernet.
As we’ve seen from our router Labs test,
you can expect to get at least four Ethernet
ports on a modern router (although mesh
router nodes tend to only have one), all of
which will be able to hit speeds of 1Gb/sec.
Network switches with many more ports are
also readily and cheaply available, making
it very easy to provide a dedicated wired
connection to as many devices as you need.
What’s more, the maximum rated length
(to still achieve full speed) for Ethernet is
100m, so in practical terms, you’ll get full
Gigabit speeds to your devices wherever
they’re located in your home.
You can even crank up the connection
speed by a notch, as many modern
motherboards come with even faster 2.5Gb/
sec Ethernet, although as our Labs test
has again proven, 2.5Gb/sec Ethernet isn’t
commonly supported by mainstream routers
yet, so you’ll need to shop around to find
devices that can deliver the extra speed.
You can also stretch to 5Gb/sec or even
10Gb/sec Ethernet. However, these two
options require more expensive, higher-
quality cabling. Gigabit and 2.5Gb/sec
Ethernet will work on Category 5e cabling,
so if you already have a Cat 5e cable networkThe ping test simply pings a reliably quick nearby server, which in our case was registered as being in Manchester
Cat 6a Ethernet cable can hit 10 Gigabit speeds
but it’s an expensive option