0
\$\begingroup\$

It seems that a power line protocol would be a really good way to help increase the range and reliability of wireless IOT networks. Wireless networks are always going to have areas that are hard to reach, even with mesh topologies, and it seems like power-line communication is an obvious choice to bridge gaps in low bandwidth networks. Yet new IOT standards like Thread have no provision for integrating power-line communications.

Is there a technical reason for this to be the case, or is wireless just a "sexier" topic these days? I notice that available standards, X10 and UPB are closed and have royalties. Is the lack of a reasonable, open standard discouraging adoption?

\$\endgroup\$
  • \$\begingroup\$ In addition to the other reasons given, there have been some high profile scam artists who promised high bandwidth data over powerlines but could not deliver. (They did get a lot of money to develop the technology, though). \$\endgroup\$ – mkeith Sep 2 '15 at 3:35
4
\$\begingroup\$

Well it's been a few years since I worked with it but at the time it was slower than wireless for one, I'm talking about the fast ethernet replacement speed parts not something simple and dead slow like X10. The good chips were controlled by pretty much one company and several patents but that looks like it's freed up now. Regardless the chips ran hot and sucked down a lot of power. Also they were limited in the amount of bandwidth they could use due to FCC conducted emissions rules, basically how much junk or modulated traffic they were allowed to stuff on the power line intentionally. At the time maybe 5-8 years ago they were lobbying the FCC to try to get more bandwidth allocated to them.

Then yes the tech was fairly closed back then so that was a little discouraging. Cost was another big factor it was cheaper to use a wifi module than a homeplug device. Of course there was a major shift from desktop to laptop, phones and tablets that all demanded wireless (hard to believe there was a time I had ethernet cords snaking over the top of my couch for our laptops :)

Now past that is a tricking little problem in the US about hopping phases. Your house get's 240 from the street and splits it at the panel into two sides of 120. Well the modulated signaling of power line communications can't really bridge that gap and there were all sorts of solutions such as having an electrician install a bridge in your fuse box... Or using wireless to hop.

But now you see it's starting to get more complicated. Manufactures don't want to saddle their wifi routers with the not insignificant cost of power line networking, so it's non standard. What you ask about range, could be fixed with multiple wifi routers working together, of course it could be fixed with two home plug devices too, but either way this is a problem for a small subset of users.

Overall it just didn't find it's place in terms of cost, performance, customer wants. There were some cases our marketing told us about that really wanted this for instance older all brick structures in Europe they said or other buildings that a wifi router had trouble penetrating, still though we had trouble justifying price for the equipment to build a powerline backed wifi network. Technically of course it could be done :)

That's the faster stuff, we also looked at exactly what you are talking about for slower speeds, and there were more options there as well although they were also hampered by the split phase problem. Then once you get outside the home it's difficult to get very far since it's hard to get through giant transformers. Inside the home what you say can be done, and is in fact done by smarthome (and probably others) next generation of X10 type equipment. They use a X10 like backbone and compliment it with lowspeed wireless to try to improve the network reliability of both.

\$\endgroup\$
  • \$\begingroup\$ That's very insightful, thank you! It seems like it has limited application, but if it could be implemented cheaply, it might be a good fallback option in a wireless mesh sensor network. \$\endgroup\$ – Obi_Kwiet Sep 2 '15 at 13:39
  • \$\begingroup\$ Yes I think so, one other thing I didn't mention is that doing a power line com design requires you to design the power supply all the way back to the plug which means UL safety testing and additional work and cost that a lot of engineers aren't used to having to do :) \$\endgroup\$ – Some Hardware Guy Sep 2 '15 at 13:41
0
\$\begingroup\$

If you are talking about wide area networks then the main reason is because in most cases the powerlines are already carrying data. The power companies won't let you transmit data on their lines because they're using it to transmit telemetry data (either to avoid paying rent to telcos or in some cases it's because it's the only option they have).

It's because of this that x10 is banned in some countries. In those countries most houses don't have isolating transformers and the x10 signals inject noise into the data stream of the power companies.

Still, there are power utility companies that offer internet access: https://en.wikipedia.org/wiki/List_of_broadband_over_power_line_deployments.

So it's not quite true that there little or no interest.

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.