You’ve heard of cognitive radio and the promise of bandwidth where the TV channels are not. We now have a standard for at least one of those. Please welcome Weightless to the list of standards for unlicensed “white space” communications. I recently heard a good presentation on what Weightless is and what its potential is.
Ed Hightower gave a presentation to the IEEE Communications and Vehicular Technology (IEEE-CVT) Dallas chapter on “M2M + TV White Space + Weightless.” He was kind enough to post his slides and audio. It’s interesting to me for several reasons. Weightless is designed specifically for machine communications. It is a low bandwidth, high latency protocol that is cellular in nature. What I mean by cellular is that there is a base station and end terminals. It is designed for low power consumption and long distances. How does it do that? You put the powerful, expensive electronics in the base station and then you can have cheaper, lower power parts in the terminals… like what is done with cellular base stations today. Base stations are supposed to be 1/100th the cost of a cellular base station and handle many more devices. The end terminal chips are supposed to be 1/10th the cost of comparable cellular modules. There is even one company, Neul, that has a chip out for end terminals. It’s high on promise.
What About the Infrastructure?
Who is going to put in the infrastructure? Since it is unlicensed you can put your own in… or wait for someone else to build a network you can purchase time on.
When can you have too much Infrastructure?
It sounds wonderful, just plunk down $10k for a base station so all my devices can talk to one another. Now the business a mile down the road does the same. Then the guy a mile from him does the same. What happens when you have more base stations than spectrum? Throughput goes way down. Most of your time is spent trying to be the loudest or retransmitting because someone stepped on you. Hopefully the error correction and spectrum hopping in the specification will alleviate most of the retransmission issues. We will see.
What happens as spectrum fills up? That’s where things get nasty. Companies will have spent millions building their network and the devices that attach to it. What happens when they can no longer count on getting the performance out of the network they need? What are the technical and regulatory mechanisms to solve contention? There are going to be other white space users and when the space starts crowding up, those companies that have invested the most into making a network will lawyer up and talk to their friends in the regulatory bodies and governments.
Long Term Prognostication
I like the Weightless standard from what I’ve seen. My concern is that long term we will see the FCC auction off the “white spaces” spectrum so that those companies who have the most to gain are able to claim what is “theirs” and make sure that it is not invaded by others. We haven’t seen this with Wi-Fi because of the short range of access points. With a six (6) mile range, there will be too much interference between neighboring base stations and terminals without some way to play nicely together… which will end up with someone claiming the spectrum in a given location as theirs because they were either their first or have the greatest need… etc. Oh, by the way, there are others who have uses for white spaces as well.
The Path to Success
I see the Weightless spec succeeding if there are carriers setup in different regions that can play well, or roam, with each other so that it is just easier for most companies and people to use the existing network than building their own. I can see cases where it makes sense for a corporate park or large industrial complex to have their own base stations. I can see the appeal of that. I’ve got some questions out about how things operate in a congested fashion, as that was not covered in the talk and the spec is not free to the public. As this “free” spectrum becomes used and valuable, expect to see turf wars among users, not just those implementing Weightless.