Sunday, December 12, 2004

Thoughts on wifi

Well, I finally managed to get my hands on one of the scarce laptops in the organization which do have Wifi. I needed it a week for a conference I went to in Washington DC. This blog-entry is a compilation of some thoughts that occured to me when using it.

  • Hooking up to public accesspoints is not always easy. Attingo's signup at Schiphol Airport is pretty straightforward, but finding the page where you could sign up with a creditcard wasn't. Notified them. No complaints further, good connection in the United Business lounge.
  • T-mobiles setup at Washington DC's Dulles Airport is a regular pain. You need to enter three secret codes: Creditcard number, user name password and something else. Pain is, that when you make a mistake it doesn't remember either. Which means you have to reeenter it again, and again and again. Whenever you forget something it forgets all. Can't they just build something that at least remembers the stuff you got right first time around?
  • STSN aka Ibahn who provide wireless at Marriotts are best to be avoided. Wireless only worked in with the hotel lobby access point, which kind of reached my room, but that was shere luck. It turned out that even though they provided conference rooms with wireless, my 9,95 didn't pay for a connection there. This according to them had to do with different security arrangements. Yeah right! specifying a v-lan isn't that hard. It was just a way for them to try and make money from the conference organizers. "if you don't buy conference access from us, your attendees will not get a W-lan connection, even if they pay 9.95 in the lobby". Oh yeah, I tried it in the hotelroom, using the high speed ethernet connection and a cable, but that didn't work either, but that just might have been me. So basically 9.95 allowed me to sit in the hotellobby, which I consider a huge pain in the behind, which is a quote they can use on their testimonials page!
  • Wireless is to be had in the area of business class/first class lounges. Wireless doesn't care if you actually are first class :-)
  • setting up secure connections is too difficult for the lay person. We need standard Diffie-Helman key exchanges. I saw on the internet that it is available on some access points, but it just should be the standard of the IEEE. As far as I could find with Google it isn't yet.
  • Securing accesspoints should be mandatory. There are too many open access points available. There is no use for anonymous connections over a random family's access point, it only endangers them into being seen as cybercriminals.
  • If people want to make it possible for neighbours and strangers to make use of their access point it should be done in the same way hotspots are now available at airports and Starbucks. Make it possible to extend the official network of the ISP to a users access point. This way if I open up my laptop and there is an access point available of Joe User, I can only hook up to it by propperly logging in to the ISP's network or use the airport/credit card system. This will require many roaming agreements etc, but it would bring security and convenience at the same time. It should be done in such a way that the person opening up his network in this way can throttle the speed of the guest users and/or the times they can access. So I would like to see a rule like "Guests can only connect when I am not connecting" or "Guests only get 1mbit/sec".

Sunday, November 28, 2004

Promiscuous wireless information

Promiscuity is generally frowned upon, and rightfully so. Before you know it you're infected with all kinds of viruses and/or is your reputation down the drain. When applied to wireless information it should be regarded as having the same problems, but that doesn't mean it should be frowned upon, actually the promiscuity I am promoting here could be beneficial to all.

What I am proposing is a way to deal openly with information, which can be openly available and shared with all. What it should accomplish is that various devices openly communicate with each other data they have available. The result is that a device can "know" more than might be possible with its physical components. A simple example is a GPS receiver. If it is equipped with a radio transmitter, it could just broadcast the current coordinates and maybe even an adress, without restricting who can be the receiver. If it did this openly other devices could use this information for their own benefit.

More and more devices get an antenna. The current way we deal with communication between devices is that we try to restrict the access others have to the information that is available from the device. Bluetooth and Wifi both try to encrypt the data and keep it hidden from anybody but those with whom a trusted a relation has been established. However much of the information is not secret, nor privacy-sensitive. Coordinates are only privacy-sensitive when they are used in conjunction with an identifier. Without an identifier a GPS is broadcasting only a fact (as it perceives it); the coordinates of here are x,y,z and the street is this. Such information could be interesting for other devices that don't have the technical possibility to receive such information directly. Now if cars already have a GPS, then why would my PDA need to be linked to a GPS I own, when it can also get that information from a car which is close by enough?

Other information I can think of that could be broadcasted in such a way is: time, weather information, news, restaurant information, sights, advertisements, coupons etc. There is so much information that people and organisations want to make available for others to freely look at and that we currently don't have a format, nor a way to communicate for. This should not result in a system that just keeps on pushing data to our devices, clogging them and making them unusable. No this would require on the one hand transmitters that transmit data (continuously or provoked) and for users a device that can be configured to say yes and no to information to know the difference between what we want and what we don't want. So I would configure my Palm to always get all the GPS data that is being broadcasted. This way I can know where I am. Information on sights and restaurants I would only enable when I really need it and am in places where I normally don't go. News I would only allow to be transmitted to my device when I enter a railwaystation, much like receiving a free newspaper at the railwaystation entrance (think Metro) . Supermarkets could inform their shoppers this way of the location of bargains or just a map of their store.

So how should this work then? I personally think it would best work with a network technology like Wifi. This because Wifi works at distances of at least 10 meters, but probably more. It is already ubiquitously available and so doesn't require new radio technology, transmitters and receivers are already available. On top of this we need an XML-like information format, which is easy for various programs to interpret. You wouldn't need a full connection to networks like the internet, because the information is already coming to you. I do foresee that people combine it with partial or full internetaccess (you get info on a touristic sight, with an URL leading to the official homepage) . All in all I feel this could work with just every device that nowadays already has Wifi built in, so it would be more a question of maybe reconfiguring some drivers, defining a standard and building software that can do something with the information.

Would this be succesful? Well maybe, maybe not. It really depends on how it is implemented, but because of the short range that the information is being send and the fact that users can define which information they want, it might be a better way then some of the proposals I have seen for location based services, which are very much push based. Plus that the short range of the transmitter makes it more localized.