VoIP over WiFi + Dysfunctional hotspot market = ?

I purchased a Nokia E65 in April primarily because I wanted to play with the WiFi and VoIP capabilities of this device. So far this had been limited to using my phone at home as a cordless phone and testing it over various connections in NZ. While I was in Australia last month I spent a fair bit of time playing with my Nokia E65 connecting via both Telstra and Azure WiFi hotspots. Connecting to my Asterisk box at home is great, I can make calls as though I’m at home and only be paying for internet charges (around A$6.60 per hour with Azure). Call quality has been perfect using both providers and VoIP over WiFi is a great solution for bypassing the high roaming costs charged by mobile networks. But is it a replacement for your cellphone? The simple answer is no.

We’ve been hearing for years that VoIP over WiFi would kill off cellphones. Anybody who still believes this needs to get a grip on reality and accept that this is nothing but a dream. With the exception of the few large scale WiFi Networks (such as The Cloud in London) WiFi networks just can’t achieve an acceptable level of coverage to ever have a chance of doing this and people don’t want to walk around a city looking for a hotspot that they can use their phone from. Pricing is a significant issue with most hotspots offerering time based charging which may be fine for web browsing but if you want to use a VoIP solution or a instant messenger application then paying hourly rates while your connection is idle can prove rather expensive. Data based charging solves this problem but has the potential to drive down the average hourly revenue generated by a hotspot. In the world of broadband however nobody ever pays their ISP hourly rates for a broadband service so why should the hotspot market be any different? Flat rate charging is possibly the solution to the problem, but more about that later.

Hotspot operators also haven’t yet overcome the most basic of problems – that 99% of hotspots use a web based interface for user authentication. Even if you are in a good coverage area trying to open up a web browser on a small cellphone or PDA screen to enter your login details is just a horrible end user experience. The solution to this is simple – use a protocol such as 802.1x EAP or WISPr to authenticate end users without them having to login manually to the hotspot. EAP simply hasn’t taken off in the hotspot world but Boingo, FON & iPass have all adopted WISPr and and any hotspot that is a partner network supports for these supports WISPr - the problem however is that most hotspots only support WISPr for their Boingo & iPass logins and not their own customer base. I have a piece of 3rd party software on my phone called SmartConnect which supports WISPr for hotspot logins however I’ve so far been unable to test this because neither Azure or Telstra support WISPr for their own user accounts and in New Zealand neither CafeNet or Telecom’s WiFi hotspots support this either. Even Steve from Tomizone has commented that he would love an application for his new Apple iPhone to automatically connect to a TomiZone Hotspot. Everybody seems to acknowledge that this is a problem but why is a solution taking so long?


Nokia have chosen to support various 802.1x EAP methods of user authentication in their S60 firmware but have so far ignored WISPr. Both iPass and Boingo have also announced Symbian S60 based client software however while iPass issued a press release over a year ago saying that a Symbian S60 client was being launched but trying to get your hands on this is another story. Boingo are currently beta testing a Symbian S60 client which will work with their Boingo Mobile service, a flat rate US$7.95 per month service that offers unlimited VoIP and internet access from a Smartphone or PDA using any of their worldwide hotspots. Looking at it the Boingo Mobile deal has the potential to finally bring VoIP over WiFi to the masses. The Netgear WiFi Skype phone has Boingo software embedded in it and Boingo is trying to establish partnerships with manufacturers to embed their hotspot software into products which allows users to simply stroll up to a Boingo hotspot and their device will automatically connect to the internet with no requirement to manually enter user details. Kodak have had Boingo support in several of their WiFi capable cameras for some time now allowing easy transfer of data from any Boingo hotspot anywhere in the world.

This leads on to the next issue – how do any of the small standalone hotspot operators who are starting up today really think they can make money? Sure, there will always be a market for a hotel or café to have it’s own private WiFi system offering internet access and many of these are probably delivering a tidy profit. As an end user however I want a single login that works when I’m at a hotspot which is exactly what large global networks such as iPass and Boingo are offering. By being an aggregator and partnering with WiFi operators around the world they no longer need to spend large amounts of money rolling out their own hotspots but relying on WiFi network operators to partner with and delivering end users to those partner networks. iPass and Boingo differ significantly in that iPass is a true roaming network where you use your existing ISP’s login to connect to hotspots (and dialup) connections any pay an hourly rate for access. Boingo charges a flat monthly fee for access to the Boingo network with some restrictions in place preventing excess downloads from single hotspots and per minute surcharges that exist on some partner networks.

I can hear people saying now that this is just rubbish but people said the same thing about roaming between GSM networks in the early days. If companies such as Boingo manage to start embedding their software in large numbers of WiFi devices then choosing to ignore partnering with large scale aggregators such as iPass or Boingo is a decision that’s going to seriously cripple your business and really does kill the whole concept of having affordable internet anywhere and everywhere.

I was going to talk about the concept of Muni WiFi networks and why they’re nothing but a bad business model but I’ll save that for another day..

http://www.geekzone.co.nz