Back in 2002 when a colleague asked ‘How do trans-Atlantic mobile calls work, because surely the signals can’t reach that far,’ our office broke up laughing. But the new invention (actually it’s a new implementation) at Qualcomm is really about mobile phones speaking directly to one another. It is calling the idea peer to peer, but that’s a misnomer since P2P networks on the internet have all sorts of ramifications, some of which this might emulate, and many of them which it would not.
Qualcomm will demonstrate what it is now calling FlashLinq at Mobile World Congress this month saying that it enables devices to discover one another automatically and then to communicate at broadband speeds without going via a base station. What’s new you ask since Bluetooth has been doing that for years, and now WiFi can do that in phones?
Well FlashLinq uses TDD licensed spectrum and is a synchronous protocol running over OFDMA carriers. So this is for operators to suddenly come up with applications where this would be relevant. The invention raises tons of questions about how such a service could be used and is limited as much by operator imagination as by application design. The bigger question is how do you charge for something like that and even before we ask that, why wouldn’t it be utter chaos if everyone tried to use it at once.
The discovery range is being touted as up to a kilometer though we’d like to know what happens if you happen to use this TDD spectrum for other applications which are base station or transmitter connected, surely there would be interference that is tough to manage except in a macro way. So if you wanted, for instance to use IP Wireless TDD broadcast transmissions for TV delivery, you would have to segment the spectrum, one part for one application and the other for P2P traffic. Most likely the TDD spectrum would have to be dedicated.
At that distance in a city you would take in thousands of phones, and the discovery capacity says Qualcomm can handle that many distinct addresses. The system claims to have distributed interference management, but by that we assume that this is a detect and avoid strategy or a simple process like WiFi where everyone that is sharing it at any point gets a fraction of the bandwidth/throughput.
Qualcomm says that FlashLinq can effectively create a “neighborhood-area network,” where fixed and mobile peer applications can interact directly. So what could this lead to? Well a cheaper form of broadband for a start, why not transmit broadband requests out on this local link, to device which can do html look ups? and return the results? Or use it like a walkie talkie, routing push to talk over an alternative data path? Or offer high bandwidth video phone access using P2P or location based advertising, which reaches out with P2P to nearby handsets with advertising offers.
All of these suggest themselves because there would be no way of compiling usage stats, so charging would have to be flat rate. If the operator sold a phone with the capability it would have to charge to switch it on, but couldn’t really track usage. Local presence apps and tracking apps could use this vehicle, to answer ‘Where are you’ questions for machine to machine or for people and pets.
FlashLinq pushes the idea of proximal communications, where users can continuously connect, disconnect and communicate directly with other mobile users at broadband speeds based on their physical proximity. We presume that the closer you are, the faster the speeds are and the more bandwidth you are allocated.
Qualcomm says that the technology is designed to complement traditional cellular-based services and serve as a scalable platform for new types of applications. Already South Korea Telecom says that it will run a trial of FlashLinq in South Korea and explore potential commercial uses.
At some stage some idiot will come up with an application that can share music from one phone to another, downloading the files, or a Skype-like app for delivering flat rate voice, but clearly that’s not quite what Qualcomm has in mind.