Yesterday I was asked to sit on a panel of 'experts' (a grouping of which I'm loathe to include myself) at the Mobile Messaging & Value Added Services conference in London. The topic of discussion was around the issues of mobile data network capacity and the potential friction between apps developers and content providers and the networks.
It was a really interesting discussion, and the research I had to do led to some interesting revelations which I want to share with you. So here are my thoughts on what might happen to the mobile data networks as usage increases, who will see the benefit in terms of revenues and who will really drive the issues.
Exposing the Myth of Network Saturation
Firstly it is, currently, a myth that the network is struggling because of the rich, high volumes of data being streamed over them. In fact it's the signalling for handling data calls that is currently causing most concerns for operators. Over 80% of traffic on the mobile data networks (I'm talking 3G here) is for handling small, frequent call attempts. These are small transactions like iPhone notifications (100bn of these were generate in the last 12 months), co-ordinate look-up from mapping applications and such like. That said, as our usage patterns change the volume of rich data we wish to transmit over the network may increase. The debate on the panel yesterday was around whether this would actually happen. Will individuals really wish to stream the latest Disney film in real-time, over the 3G network whilst they're sat on a train? Probably not and they'll continue to download such rich media at home over wifi and fixed line broadband prior to their journey, storing it on the device for later viewing. It's probably new products such as the Apple iCloud that could eventually grow our hunger for rich media up- and down-loads as we dump everything into the cloud and it is then all pushed to our numerous devices...
Bypassing the Network Bottleneck
For now though, it's signalling that's the issue and the immediate concern. The problem here is that the RNC is a bottle-neck in the network. This isn't being helped by operators sharing infrastructure as it just increases the load on the carrier's RNC. It could help by installing new masts and cell sites, but this is difficult for planning and cost reasons. You can duplicate the number of RNCs to improve their call handling capacity, but again this is an expensive and not necessarily the final solution in any case. What is needed is bypassing the RNC for certain types of data. By breaking out of the network before we reach the RNC and into other IP networks to carry rich data, we can free up the RNCs to deal with the small, processor hungry data sessions such as notifications.
If we can push more data onto wifi, wimax or femtocells (especially metro-femtos) we can handle rich data much more efficiently.
Then we can make use of that brilliant network infrastructure the SMSC. Why not use SMS instead of data notifications where possible. There's plenty of capacity on the SMS network.
Breaking out, offloading or sideloading (whatever you prefer) is great but it needs intelligence in the network and on the device (or in the app software) to seamlessly seek out, connect with and transmit over suitable alternative data networks. Same for SMS, it needs apps developers not to be lazy and use notifications because the API is easy to access and to make some forward thinking decisions about what can be carried as an SMS.
Obviously the the other limitation of off loading is that the wifi/wimax infrastructure simply doesn't exist with sufficient ubiquitous coverage to make this effortless. Also, the charging mechanism are barriers to take up - who wants to pay for wifi in coffee shops when they're already paying a bucket of cash for fixed line broadband at home and 3G data plans on their mobiles? BT should be applauded for addressing this issue, shame they can't fix the network coverage at the same time...
The other major barrier for off loading to overcome is that of control. If the operators let valuable transactions leave their network before they get hold of them in the core, they lose control of how they're transmitted - over who's network but also they lose the chance to add handling fees on to the billions of transactions we expect to be generating over the next few years (mobile payments for example)... It's a numbers game and someone needs to take the bull by the horn and sort it out - the regulator perhaps...
Get Rid of Unnecessary Data Traffic
Whilst we're having a go at operators let's dig the knife in even more... How can an operator complain that their network is becoming full when they subsidise devices that consume huge amounts of data? You can't have that cake and eat it. What about this strategy of competing directly with the fixed line broadband providers. I'm not talking about O2 taking advantage of OFTEL's obsession with local loop unbundling and offering ADSL services in the home. I'm talking about 3G dongles that are advertised as an alternative to fixed line broadband as opposed to a useful way of getting your email when you're not near a wifi hotspot... Don't sell capacity that you don't have. If you've got plenty of wool but not much fabric, stick to the knitting and don't try to sew a jacket...
What Can Content Providers and Apps Developers Do?
I'm not going to push all the blame to the network operators. Us apps developers have a role to play because, like it or not, if the data network becomes congested and service levels decline or, heaven forbid, the networks try to exclude and discourage data usage through high cost data plans, then we won't be able to sell so many apps...
Apple are playing a dangerous but exciting game. They've decided to go it alone and make cloud technology the norm. The launch of iCloud and the fact that it's free is a nightmare scenario for operators worried about network capacity. Let's face it, they're not going to win a war against the might of Apple, especially with giants Google and Microsoft about to join in as well... Apple want to sell more of their devices (they're also targeting the 30% of households in Europe without computers) and they see the new LION OS, iOS5 and iCloud services being the drivers for these sales. The fact that these will be network hungry solutions doesn't worry them because they know the networks will fight each other to stock the next iPhone 5, the iPad 3 and so on... Network Operators will face a dilemma of selling the phones that people actually want, or selling the tat that they don't want and protecting their network... Apple will win...
One good piece of news for the operators coming out of the iCloud product launch announcement though was that there will be an API to allow apps developers to access and configure the intelligent routing capability of the iCloud - so timely downloads using wifi etc will be seamless to the user.. Hooray!
As apps developers we need to start thinking about network capacity. Same as for oil consumption, if we use it all up to fulfil our immediate desire to hare around in Ferraris at 180mph, in 20, 30 or 50 years time we'll be forced to drive Toyota Priuses whether we want to or not. Same for apps - ignore the future issue of capacity now and you'll be consuming your own arms...
To summarise. Mobile Data Capacity. It's more about signalling than rich data today, but rich data will become a problem as well. The network operators need to sacrifice some of their huge margins for the good of mobility. Maybe the regulator has a role to play as it did with BT and the fixed line broadband operators. We all need to get a piece of the pie.
0 comments:
Post a Comment