The first International Symposium on Applied Sciences in Biomedical and Communication Technologies (ISABEL 2008) will be held from October 25-28, 2008 in Aalborg, Denmark. (Note: it overlaps with the SDR Forum). The call for papers is here (pdf). Relevant dates:
- July 25, 2008 Submission deadline
- September 10, 2008 Notification
- September 25, 2008 Camera Ready
Relevant topics (in addition to the more biomedical specific topics) include:
- Game Theory for Flexible Spectrum Usage
- Optimization Problems for Cognitive Radio and Advanced Spectrum Management
- Bio-Inspired Algorithms for Information and Communication Technologies
- Machine Learning for Cognitive Radio
Voyant got a contract to build white space radios. 3000 doesn’t sound like a real big order. But $2 million doesn’t sound like a lot of funds for development either.
Larry Page is in DC today to talk about google’s white space proposal with Michale Calabrese. (via) I think the whole talk will be available at that link. Listening, I’m hearing lots of good things (support for dynamic spectrum markets) and lots of objectionable things (e.g., an all frequency, all-mode SDR will be $5).
(Quasi live-blogging – ignoring the stuff not directly related to cognitive radio / wireless)
Larry asserted that you could turn off a transmission in a single ms. However, packet/frame lengths and processing times for the high data rate systems they want to use simply won’t support that.
Lengthy comment in support of completely turning off TV broadcasts and switching everyone to cable/satellite, but noting political difficulties (particularly must-carry requirements). I think that was about right.
Discussion of drop in relative ranking of the US broadband access. After noting that some places in the US have very high broadband penetration (like DC), Larry went on to lament the lower access levels in the US. My thoughts are it’s primarily a function of the lower relative urbanization of the US which makes deploying broadband far more expensive + the US deploying broadband earlier (and now stuck with a bit of legacy equipment).
Larry defends the business model of municipal WiFi?
I decided I was being a bit too negative in tone on a couple things in the post. I think this was because of my recent disposition to the two subjects in the blog post. I need to read it again, but I think I have a significant theoretical issue with the Google white space white paper, and the Voyant news came via a stock advertisement which looks the world to be flogging a dying stock.
So tempering a couple of my negative comments:
1) First, while various standards have long frame lengths and just because it normally takes a while to sense a signal, it doesn’t mean that a different standard couldn’t be defined that permit that level of responsiveness. However, I still expect this will severely cut into the achievable data rates (you lose transmission times istening, plus lose efficiency by using shorter frames)
2) A $2 million dollar order would be a great order for a startup. However, Voyant is not a startup. Nonetheless, this is the first white space device order I’ve seen, so that’s good news for the cognitive radio community. However, it seems just a bit premature as it’s far from certain at this time that the regulatory environment will permit their use.
The FCC has issued an Notice of Proposed Rule Making to revise the operation of Block D (pdf).
The key sentence for me:
First, we seek comment on whether it remains in the public interest to require a public/private partnership between the nationwide D Block licensee and the Public Safety Broadband Licensee for the purpose of creating a nationwide, interoperable broadband network for both commercial and public safety network services.
A lot of the other key issues for which they are also seeking comments (negotiating rates, setting auction price, network sharing agreement) becomes inconsequential if instead of looking to create a two-headed monster the Feds create a single government sponsored enterprise chartered to serve the joint mission.
The FCC is also considering just abandoning the public safety communications provisioning requirement, which I think would be a mistake.
On the Google public policy blog, there’s a link to a Google white paper (pdf) describing their geolocation proposal in more detail (though not a lot more). As opposed to earlier reports, they appear to be deemphasizing the aspect of having existing mics upgraded to include a beacon and are going more for a pure geolocation / database method (I think that’s preferable).
They’re also engaging in a little bit of the “cure for everything” mentality that is an unfortunate trait of the cognitive radio community. Specifically, Google is also touting the ability of unlicensed mesh White Space devices to aid the public safety community after a natural disaster. While the combination of mesh networks and cognitive radio is frequently touted for those purposes, I think it’s an unneccessary distraction in this case.
Specifically, what the unlicensed White Space community needs to focus on is just demonstrating non-interference with incumbents – which unfortunately hasn’t gone as smoothly as it could have. As additional features creep into these prototypes, the chances for bugs will neccessarily go up and any error at this point is being seized upon by the incumbents as proof of the infeasibility of DSA & cognitive radio.
In the last post on Google’s white space proposal, I mentioned that theaters and churches were also objecting, but I did not have links to support that assertion. Finally, here’s some news articles to support those claims.
Grand Ole Opry (along with CMT and MTV) complains about White Space.
A church audio blog following the controversy. It’s a short link roundup, but the Shure link is interesting to follow (they’re against it). While I’m on the topic of Shure, Michael Marcus rightfully rips into them in this post where he points out that frequencies suggested for use by Shure for wireless microphones, often unlicensed, interfere with licensed public safety incumbents! (Someone once said something about motes, beams, and eyes that seems apropos.)
Management of Vertical Handoffs
A news release noting an installation of a cognitive system from In Motion Technology for East Texas Medical Center (ETMC) Regional Healthcare System Emergency Medical Services. Reading into their solution page and technology page, what they’re actually doing is optimizing the choice of network assignment based on link information and performing vertical handoffs. Think normal mobile assisted handoffs (maintain a ranked list of link state information from multiple base stations and switch when some hysteresis threshold is crossed), but not restricted to a single wireless interface.
It’s not what most people would call full-blown cognitive radio, but this is representative of how I think cognitive radio is going to roll out – function by function.
White Space Device Resubmission
Motorola is resubmitting their white space device to the FCC for testing. Not much information in there but it’s interesting to note.
US Navy Video on spectrum
Via Elastic Spectrum, comes this video entitled, “Electromagnetic Spectrum, Critical to our Nation’s Security and Economy.” Beyond the video (kind of an overview), there’s also several nice reference documents linked on the sidebar at the site
Primarily, the article summarizes contract awards in the development of SDR and cognitive radio.
1) There’s a claim in the article that cognitive radio will be cheaper than a normal radio. While I expect that to eventually be true, I don’t see it happening in the first few generations of cognitive radio. Basically, there’s going to need to be changes in the way radios are designed, the way systems are specified, and the way spectrum policy is specified. And all of that is not going to happen at once.
2) Also on the subject of price, the $500/radio DARPA quote is just for the brick (hardware). Software development will significantly add to the price tag. As radios transition to SDR, the majority of the cost will increasingly come from the software. This is to be expected as the goal of SDR is to realize as much functionality in software as possible.
3) The article mentions the value that IT saw from open standards in terms of decreasing prices. Drawing an analogy with IT, I also expect the value from SDR and CR to be minimal, if not more expensive, when just replacing existing solutions. Real value will come from new applications and as processes are adapted to leverage SDR and CR.
4) Bitwave has made big claims for their RFICs which I’ve cited in presentations. However, I’ve heard of delivery and performance issues. Anyone have any first-hand information?
Got an email announcing the Wireless Internet Conference (WICON) to be held from November 17-19 in Maui. In its Emerging Technologies and Applications track, a subtopic is cognitive radio.
Other areas of interest include: Seamless Integration of Heterogeneous Networks, Cross-layer Design and Optimization, Wireless Access Technologies, Multi-hop Wireless Networks, Network Security, Wireless Internet Platforms and Software
Key submission dates:
- Submission deadline: July 15, 2008
- Notification of acceptance: September 15, 2008
- Camera-ready version:October 15, 2008
Via an email comes an announcement that there will be a Eurasip issue on Game Theory in Signal Processing and Communications.
- Static non-cooperative games (Nash and Stackelberg equilibria)
- Finite and infinite dynamic games
- Cooperative (bargaining) game theory
- Auctions, coalitions, and pricing
- Game theory for resource allocation in communications
- Game theory for adaptive waveform design
- Game theory for cognitive radio and dynamic spectrum access
- Stochastic games, repeated games, and fading channels
- Development of decentralized algorithms using game theory
- Manuscript Due: October 1, 2008
- First Round of Reviews: January 1, 2009
- Publication Date: April 1, 2009
I’ll submit a paper, the only question is on what topic. So readers (some of whom I think are familiar with my work), which of the following would you prefer:
In Reuters earlier this week, the NBA, NFL, MLB, and NHL collectively issued a statement as the Sports Technology Alliance in opposition to the use of the White Space (unoccupied TV bands), and particularly singled out Google’s plan.
We are deeply troubled by the very serious disruption and harm that portable device interference will cause to sport broadcast programming, whether prerecorded or live, and the conduct of the games themselves,” the Sports Technology Alliance, said in a filing with the Federal Communications Commission.
“We are deeply troubled by the very serious disruption and harm that portable device interference will cause to sport broadcast programming, whether prerecorded or live, and the conduct of the games themselves,” the Sports Technology Alliance, said in a filing with the Federal Communications Commission.”
Basically, they think that the wireless mics will be excessively interfered with. Google responded by basically calling the Sports Technology Alliance a bunch of Luddites (probably not too helpful).
“It’s really too bad that some people prefer the comfort of the past to the promise of the future, and defend the status quo instead of working to bring the Internet to more Americans.
“We enjoy the Super Bowl and NCAA tournament as much as the next sports fan, and wouldn’t support any plan that interferes with professional sports”
Similar thoughts have been expressed by hospitals, theaters, and churches (sorry, can’t find links for the last two). All of these make some use of the TV spectrum now (wireless mics for the latter two, medical telmetry in Channel 37 for the hospitals), and they don’t want their implicitly-government-provided ox gored.
My two cents. After having a little time to think it over, I think the Google plan (geolocation + database, + beacon for mics) can be improved to handle these complaints in three simple steps.
1) Extend the database to include churches, theaters, hospitals, and sporting venues. They don’t move, so this should be easy enough to add.
2) Add to the database location entries a field (or fields) to capture permissible transmit times. Most of the time an arena, theater or church are not in use, but when they are, it’s not typically something the venues want to keep secret.
3) The protected regions around these venues should be much smaller than for TV footprints.
With these in place, the white space device gets its location information (either directly or indirectly), the time of day (directly or indirectly), logs onto the net (again, either directly or indirectly) and finds out if it’s ok to transmit in its current location and for how long (alternately it can periodically poll the database to make certain it’s ok to transmit).
Hospitals, churches, arenas, and theaters while they’re in use (for hospitals, this means all the time). Further I see no reason why mics in those locations would need to add the special beacon, so all of these guys could continue using their equipment as before. Of course, it would be incumbent on the white space coalition to build / maintain a good working database, but that sounds like something Google would be good at. Plus it would give an excuse to create “Google church” (find out what’s going on at the churches in your neighborhood and when), “Google Theater”, “Google Sporting Events”, and the like.
As an added bonus, with all that venue information and the Google brand, Google could also start to sell event tickets online and break the TicketMaster monopoly.
is here. I note that Google and I are on the same page.
For the TV broadcasters, Google suggested that white-space devices be required to incorporate both Internet access and some type of geo-location technology, such as GPS. That way, the devices could check an online database, maintained either by the FCC or a third party, of licensed users of the spectrum at a particular location at a particular time.
My rationale: There’s no need to bite off the whole enchilada in one go. We can quickly get someting up and running and demonstrate feasibility / build confidence for the concept of secondary spectrum use. Later we can begin deploying ad-hoc networks as the technology matures.
Note that in an ad-hoc network of limited size, if at least one device has net access and geolocation capability (caveat being there’s many many more ways to do geolocation than just GPS), then we can still have networks of mostly cheap radios.
I would also note that there should be two different sets of regulations here, one for when the CR knows its location (higher power) and one when the CR does not know its location (lower power).