What are the requirements for a Home Network?

Setting up a computer network in your home is not rocket science but it pays to think about what you want/need before you start buying equipment or drilling holes all over your house.

Often people don’t actually know what they want or need because the whole concept of a “networked home” is alien to them. So, let me give you a bit of a network primer before we get down to the nitty gritty of actually planning and installing a home network: Home Networking, a primer for the different technologies

As mentioned in the Network Primer above, a wired network is currently the only robust means of moving large volumes of data around a network at decent speeds. For streaming, whether it be audio streaming or video streaming, you must have a constant, quality connection to maintain a constant stream of data. If you don’t have this you’ll experience drop-outs or stuttering in the audio/video stream.

Is it difficult to install a wired network?

Installing a wired network in your home is really not that tricky from a conceptual point of view but you must think carefully about your requirements before you even think about getting the drill out. Whilst you could call in the experts and get everything installed professionally I personally think the job is something that can be carried out by the competent DIYer. The job basically involves drilling into walls, chasing channels and lifting floorboards so if you’ve never done any of this, successfully, before and don’t feel confident doing so, then don’t. The last thing you want to do is drill into a buried water pipe or a gas pipe or even a live electric cable! You have been warned!!

For my network I set about considering the requirements for each room individually. If you try and think about the network as a whole then you’ll likely get yourself all confused and end up going around in circles.
Once I’d considered each room I could then think about how to get Ethernet cables into those rooms. Once you’ve done that you can then think about the hardware you’ll need to buy to accomplish your goal. The hardware consists of network outlet faceplates and back boxes, cables, network switches, patch panels, RJ45 connectors, patch leads, a crimping tool, network tester etc.

I’ve heard the term “Node Zero” but what is it?

In a traditional wired home network you should establish a “Node Zero”. A Node Zero is a location somewhere in your house where all the network cables which will run around your house come together. Try to find a location which is convenient for cable runs. Some locations, such as under the stairs, are better suited than others. You need to ensure this location is well ventilated; servers and NAS devices will live longer if you provide adequate ventilation for them. The location should ideally provide access for an incoming telephone line as well as a TV feed (terrestrial/cable/satellite). It should also be free from dampness and excessive heat.

So, do I have to have a Node Zero?

If you’re building a house from scratch then establishing a Node Zero is something that is a little easier than if your home is already built and decorated. Mine was the latter so I actually have more than one Node Zero. In my house, all the network cables for the upstairs rooms run into a single 24-port gigabit switch located in the loft. My server and NASes live in an upstairs cupboard and are connected to this switch. The main living room and cinema room, both downstairs, are also connected to the switch in the loft. The reason these two downstairs rooms are connected to the “main” switch is because the bandwidth requirements for these two rooms are much higher than the other rooms in my house. Plus it was easier to run the cables from the loft into these two rooms!

All other downstairs rooms are connected to another 24-port gigabit switch located under the stairs. My broadband router is connected to the under-stairs switch to provide internet access, wireless access plus it also handles the DHCP duties (ie. it provides IP addresses for all my network devices) and I have a single Ethernet cable connecting the two switches together.

How many network cables should I run to each room?

I seen this question asked many times and I’ve seen many different answers to it. That’s because there really is no correct answer. Firstly, it depends on the room; the network requirements for a typical bathroom differ greatly to the network requirements for a typical study or your main living room. Secondly, it depends on how easy it is to install network cables in that room; If it’s a simple job to install cables without disturbing the décor in that room then you should install as many cables as you need today knowing that you can easily add more cables in the future as your requirements grow.

In my setup I aimed to install two network outlets by each mains socket in each room. The reasoning behind this thinking is simply that most network devices require power and so will be plugged into the nearest mains socket. They will obviously also require a network connection. So, the most logical place for the network outlet is by the mains socket. Not exactly rocket science really, is it!

The reason I aimed to install two network outlets by each mains socket, and not just one, is simply for “future use”. Some people choose to also run a couple of spare cables (thus making 4 cables per outlet) and to leave them un-terminated behind the faceplate. I didn’t do that. I figured I could use one of the two cables as a “guide cable” to pull extra cables through in the future. For some runs I pulled a piece of string along with the Ethernet cables and will use that string if I need to pull extra cables in the future.
Since no one knows what the future holds don’t fret over your choice too much. All I can say is to just install more than you need for the foreseeable future and leave it at that.

In the main rooms (living room, cinema room, study) I installed more than two by each socket since my needs today exceeded 2. If in the future I need more I can always plug a network switch into one of the outlets to give myself a few extra “ports”.

What sort of network switch should I buy? Do I need a “managed” switch?

When it comes to network switches there really is no particular brand which stands head and shoulders above the rest in terms of performance and/or build quality, although that obviously doesn’t stop everyone having their personal preferences.

A “managed” switch (a switch having configurable options) is more suited to commercial use than home use and so unless you have very good reason to get a managed switch (in which case you won’t be reading this page in the first place) then you should go for an unmanaged switch. Unmanaged switches are simple plug and play devices and are generally much cheaper than managed switches.

So, having decided on an unmanaged switch the other thing to decide is whether to go for a Fast Ethernet switch (100Mb/s) or a gigabit switch (1000Mb/s).
Most ADSL “routers” come with several Ethernet ports built in. More often than not these ports are Fast Ethernet ports (well, this is how life was when I put this page together). So, all devices plugged into these ports can communicate with each other at speeds of up to 100Mb/s. However, if you plug a gigabit switch into one of these ports and then plug all network devices into the switch (and not the “router”) then the devices can all communicate with each other at speeds of up to 1000Mb/s. Assuming of course that these devices are fitted with gigabit network interface cards.

So, if you intend streaming lots of HD video around your network then you’ll have plenty more headroom if you buy a gigabit switch rather than a fast Ethernet switch. They do tend to be more expensive but the extra expense is well worth it. 

There are UTP, STP, FTP, Solid & Stranded Cables. So what sort do I need?

Ethernet cable is made up of four twisted pairs of cables (so 8 cables in total) inside a single plastic jacket. Each of the 8 cables has a plastic insulator around it.

UTP = Unshielded Twisted Pair
The twisted pairs are not shielded. This is the most common type of cable used in a home network environment.

STP = Shielded Twisted Pair. FTP = Foiled Twisted Pair.
With these two types the twisted pairs have a metal shielding around them in an attempt to prevent electromagnetic interference (EMI). These cable types are typically used in those commercial environments where EMI is an issue. They are bulkier and more expensive than UTP cable and, if not installed correctly, actually provide zero benefit over UTP cable. It’s rare to find the need to use shielded cable in a home network environment.

Solid cable. Each one of the 8 cables is made with a solid piece of copper. i.e. there are 8 solid copper wires inside a single Ethernet cable. They are more fragile than stranded cable and do not appreciate being handled regularly as they are more likely to break or become seriously weakened.

Stranded cable. Each one of the 8 cables is made up of many individual strands of copper. Stranded cable is more flexible than solid cable but is more expensive.

The general rule of thumb for solid verses stranded cable is to use solid cable for “building wiring”, that is the wiring which is normally installed inside walls and under floors and connects a network switch (or patch panel) to a wall socket, and to use stranded cable for “patch leads”. That is, the leads which connect a computer/router/streamer etc. to a wall socket.

I’ve noticed there are cat5, cat5e and cat6 cables. What is the difference?

These denote the “standards” that the cables are constructed to. In very simple terms (so not 100% accurate but close enough!) cat5 is designed to operate at “Fast Ethernet” speeds, i.e. 100Mb/s. cat5e is an enhanced version of cat5 and is designed to operate up to gigabit speeds, so 1,000Mb/s. Cat6 is designed to operate up to 10gigabit speeds, so 10,000Mb/s.

If you’re installing a network today then forget about using cat5. The choice will be between cat5e and cat6 cable.

Should I use cat5e cable or cat6 cable in my home network?

That’s a tricky one, and you’ll likely read many conflicting opinions. Again, there is no correct answer to this question.

It is more difficult to install a cat6 network in your home than it is a cat5e network. There are two main reasons for this: The first is the requirements for a cat6 installation are more stringent than a cat5e installation. Secondly the cable itself is more difficult to work with. It is thicker and less flexible than cat5e and the cable is generally regarded as being more “delicate”. The cable itself is more expensive than cat5e cable plus you obviously need to use cat6 connectors, face plates and patch panels all round.
Furthermore, if you do not install a cat6 network correctly then you will be lucky if your installation even meets the cat5e standard.

So, what should you use in your network? Well, I’ve used cat5e in my home for the simple reason that it is more than adequate for my needs today. Streaming blu-ray video across a network consumes around 50Mb/s of bandwidth. Cat5e is more than plenty, remember it can run up to 1000Mb/s. I wouldn’t be surprised if “tomorrow’s needs” required a fibre optic network so even cat6 won’t be up to the job.

I’ve seen the terms TIA/EIA-568B and T568A & T568B. What’s that all about?

TIA/EIA-568B is a standard for commercial building cabling. The thing we need to be concerned about in this particular standard is the specification for cable termination. That is, how we connect the 8 wires within an Ethernet cable to a patch panel or RJ45 socket or face plate. There are two specifications: T568A & T568B. Just to make matters confusing the TIA/EIA-568B standard specifies that cables should be terminated using the T568A pin/pair assignments. However most people follow T568B. It really doesn’t matter which one you follow as long as you stick to one specification or the other. In my network I’ve used T568B throughout.

Patch panels and RJ45 face plates are normally colour-coded with both standards so you don’t have to remember the above diagrams, simply remain consistent in whichever wiring standard you choose. It’s only when you come to wiring an RJ45 connector for example that you’re “on your own” and have to follow the above.

Should I use a patch panel?

Yet another question with no correct answer! In my opinion a patch panel (also called a patch bay) in a home network environment is unnecessary, although I must confess it does look very cool!
A patch panel, for those of you who don’t know, is a panel which is often rack-mounted and allows you to plug patch leads (short, stranded-core Ethernet cables) into the front side of the panel and the back side of the panel holds the longer, permanent (solid-core, “building wiring”) Ethernet cables. The patch leads are used to create “circuits”. For example, to connect an RJ45 faceplate (the cable for which would be connected to the back of the patch panel) to a network switch.

There is absolutely nothing wrong with terminating one end of a “building wiring” Ethernet cable with an RJ45 connector and plugging it directly into a network switch, thus eliminating the need for a patch panel altogether. HOWEVER, “standard” RJ45 connectors are designed for stranded-core Ethernet cables and are not suitable for use with solid-core cable. You need to buy RJ45 connectors which are specifically designed for solid core cable. As mentioned above, solid core cables are not really designed to be moved around a lot so only plug them directly into the switch if you don’t plan on moving them afterwards. For info I have used this method in my setup and have had no issues.