Home network technologies for the next 5 years?

[infotime]

What do you think about the future of wireless networks in our homes?

Not a week goes by where I'm not reading about some other device or application that relies on home networking. Today for example, my sump pump was unplugged. Fortunately we noticed in time. Started looking at failsafe sump pump systems and find devices that will text you when there's a problem. But I'd like to focus this thread on the infrastructure in the home.

If you're going to do all these things that require a good network, then how do you make a good network? And what's the business opportunity for someone like me who does computer work for residential and small business customers.

I've worked in the computer industry my entire adult life, for the last 28 years. Since 2004 I've been on my own (self-employed, one man show) doing home and small business work. It's all been general IT type stuff - virus removal, new computer setups, WiFi setup, application help, shared printers, and in recent years help with iPhone and iPads and syncing help.

The trend is away from PCs and I'm seeing less dependence on getting home computers fixed, but I'm staying busy anyway. I *think* my long term opportunity is to focus more on home networks.

I've set up a few different systems like multi-router systems using WRT54G routers some configured as access points, Open-Mesh, Apple Airport multi-device networks and one or two Ubiquiti brand systems.

My theory is that the best networks for large homes need a wired backbone. I don't do the wiring and it's either in place already or I recommend a homeowner get their own electrician or contractor to handle that.

So in planning for my future I'm wondering if maybe I should offer a turn key package. Or just focus on that aspect more.

The main question: do any of you forsee any technology risk in building a business that focuses on hard-wired multi-access point networks in medium to large homes? In 5 years will there be a WiFi technology that doesn't need all those access points?

It seems to me that most people would prefer to buy one box and be done. That's what they're doing now, some with poor results. Maybe in 5 years that will be the best solution due to advancements in radio electronics and chip and antenna improvements.

But if it's not then what's going to be the best long term way to build good networks for large to medium homes?

 

[stevech]

The main question: do any of you forsee any technology risk in building a business that focuses on hard-wired multi-access point networks in medium to large homes? In 5 years will there be a WiFi technology that doesn't need all those access points?

My two-cents:

• 2.4GHz band will decline in popularity due to better alternatives with less shared-use competition.
• Today, price competition causes WiFi gear to not transmit at higher power even under current regulations. Especially on the handheld device side.
• Much higher bit rates = shorter transmitter on-time = lower battery consumption. But high bit rates lead to more costly hardware and wider channel bandwidth
• Curiously, handheld mobile devices intrinsically need much lower bandwidth. The more mobility, the less bandwidth needed. Human factors.
• Cracking the state of the art 2-3 bits/second/Hz of channel bandwidth isn't expected soon. The laws of physics prevail.
• Perhaps greedy cellular companies will see a way to lower consumers' cost for pico-cells in the residences
• The above may lead to less use of privately owned wireless WiFi
• Will US (et al) citizens become aware of, and fight against government auctions of spectrum licenses? $Billions of income lost in GSA. $Billions of cost borne by the cellular companies bidding in the auctions, and these costs are passed on to us. Will the US EPA start auctioning the air we breathe?

 

[thiggins]

Good question. The trend is toward 802.11ac, which requires 5 GHz. 5 GHz signals are attenuated more quickly then 2.4 GHz, so coverage per AP is less. This argues for more APs, especially for multi-level homes.

Two technologies reduce the need for Ethernet wired backbone: MoCA (via coax) and HomePlug (via electrical wiring). DirecTV and Dish already use a form of MoCA to connect a central DVR to cheaper boxes that can pull content from the DVR. It is also used in Hospitality and Multi-unit housing. It's not really a consumer technology due to installation complexity.

HomePlug is the dark horse. Bandwith has been improving with successive generations. AV500 SISO can produce 80-90 Mbps. AV1000 (MIMO) should add
at least a few 10's of Mbps onto that. The technology is here to produce small wall-pluggable HomePlug connected AC APs. But manufacturers don't see the market, at least not in the US.

But Powerline throughput can be reduced by powerline noise and other factors. Still, it can produce higher and more reliable throughput than wireless extenders.

But for the highest throughput, least troublesome networks, multiple APs connected via Ethernet will be the way to go.

 

[azazel1024]

5 years out, maybe. 802.11ax is supposedly 3-5 years out (just like everything "coming soon") and is supposed to cram 3-4x the bandwidth in. Whether it requires wider channels to do this, or manages a larger character set, like 512QAM or something, no idea. But in half a decades time, 11ac might "go the way" of 11n.

I'll just be happy when 11ac is mainstream.

Powerline is getting better and MoCA is a resonable alternative, but the problem for both is, that like wifi, they are shared medium and half-duplex. So even if you can manage 100Mbps over MoCA, its divided among all devices. Its fine for general internet access and streaming from the internet or a LAN server...but it isn't really suitable for any high speed LAN or internet services.

Powerline pretty much the same, except it can be very variable on its performance. The big perk on both of them is, in general, your neighbors can't stomp all over it like they can your wifi. They also have no problems with things like masonry or poured concrete and rebar.

However, if you want really high speed networks, you have to have Wifi or better yet Ethernet. A good 802.11ac connection can easily push over 300-400Mbps half-duplex, which is easily 3-4 faster than MoCA or a really good powerline connection, even if its a shared medium. Ethernet, for basic stuff is now all pretty much gigabit, and NOT a shared medium. So you can get 1000Mbps...call it 900Mbps or so with overhead, to AND from every single device, since it is full duplex.

Ethernet still rules supreme and probably always will. Likely by the time 802.11ax comes along and finally starts putting gigabit ethernet to shame, 10 gigabit ethernet will be relatively common on modestly high end devices.

 

[sfx2000]

I think in the next 5 years...

1) Internet of Things - smart thermostats (already here in limited deployment), smarter settop boxes, home security and alarming, better energy usage, personal telematics

2) More IP networking in the house - some of it will be self managed, but more of it will be managed outside "in the cloud" - see above

3) Wireless - 2.4GHz is still strong, not just for WiFi, but for BlueTooth EDR and LE modes and ANT - we'll continue to see growth in 802.11ac, but 802.11ad will be there as well - and perhaps 802.11 in the 900MHz space for low-bit rate things like appliances and what not.

4) The Home Cloud - it'll be about the Internet of Things, along with your handsets and laptops/desktops - and it'll be largely managed by someone upstream, aggregating the data feeds...

5) transition from ISP's to ASP's as primary data providers - things will move up the software/application stack, and those Application Service Provider will be there. This will be from "Connected Car" to "Smart Home" to the "Internet of Things".

Just a few things off the cuff...