O ye technical types (I can’t think of a better way to ask a bunch of sane technical people a hairy question than blogging it) how hard would it be to cram a high-bitrate, bi-directional signal into an ordinary stretch of train track? The length of said track could be tens of kilometers (metropolitan), or hundreds (rural, interstate)
The thing is, it would be truly sweet to have viable wi-fi broadband on the train, be it metropolitan commuter trains, or rural ones. To do this, one first needs to build a nice stable data pipe between a ground-based point-of-presence, and one or more big heavy chunks of rolling-stock per train which periodically vanish under the earth (or at least dive into deep trenches). After brief thought, the obvious answer is that there’s already a colossal pile of pre-existing cabling in the form of the tracks and overhead power cables (where electrified). Indeed, this is already used for signalling purposes and ‘train detection’.
The question is: can you cram 100Mbps (say) over it without lots of (expensive) new relays and repeaters? Can you do so without screwing up the existing system? etc.
Any insights/suggestions/links, anyone?
And then you could use Ruby and call it “Rails on Rails”.
η
Just found a site for the crowd who are wifi-ing the BART, but can’t copy and paste the link right now due to the limitations of my phone browser.
Thank you Jiri. 🙂 I should have taken a bet on your saying exactly this, if I could find anyone to bet against, and if I weren’t worried that Korny would do it first.
Very nice idea. But its not something that will ever happen. Look at it from a business perspective:
A> Pour money into a project to provide WiFi for train commuters, or
B> Pour that same money into improving your existing 3G/HSPA network, which isnt just limited to train commuters.
Kewl idea, hideously impractical 🙁
That URL? wifirail.net
The key reasons why I would see this as viable:
a) I wouldn’t expect a Telco to be interested. This is a way for a train service company like Connex to draw commuters from buses, trams, etc. and perhaps extract some extra cash from them (since the usual fiscal rules of captive wifi audiences apply).
b) The potential bandwidth of a train network is much better than that of roaming broadband: You can use a big fat terrestrial link and short-range wifi. Roaming broadband is getting faster, but so is terrestrial networking and wifi.
c) Reliable cell-to-cell roaming at ~80kph for a whole trainfull of commuters, let alone dozens of such trains, without the dropouts that make, e.g. streaming video, infeasible, isn’t something that any technology I’m aware of can currently offer.
d) It is almost infinitely cheaper to hook up a few captive APs plus some kind of fixed link for each train than to build (or augment) an entire roaming mobile network.
I’d expect factors like higher speed, greater reliability, lower cost and the fact that it uses pre-existing devices (802.11x wifi) to make it a trivial, hands-down win over expensive, slow, unreliable (when moving fast in a metal box with lots of other users, sometimes underground) wireless broadband which necessitates a new phone and much software fiddling.
Disappointingly, wifirail aren’t using the actual rail infrastructure, making them less of an automatic win over 3G and such, so my question stands: Big long heavy steel rails for moving IP traffic? Yea? Nay?
Maybe do some testing, research it, and try and sell the idea to Connex 🙂