Remember those news reports on the difficulties experienced by AT&T's 3G network due to the explosion of data generated by all those Apple iPhone smartphones? Blogger Paul Mah sat down with an industry veteran to talk about the future of mobile data and the key challenge faced by mobile operators.
I had the opportunity last week to sit down with Barry Hill, Vice President of Sales and Marketing of Santa Clara-based Stoke, Inc., to talk about the future of mobile providers. Stoke is a maker of carrier hardware while Barry has been involved in all aspects of the telecommunications business for the past 17 years at a managerial level.
It was by sheer coincidence that I brought up the topic of data consumption by smartphones, based on much earlier reports of complaints over the data consumption of Apple iPhone users. In fact, The New York Times actually ran an updated piece — which I did not see until later — of how customers are angered as iPhones overload AT&T 3G network.
So beyond the obvious fact that iPhone users consume way more data than customers on other smartphones, where are we heading on the mobile data front?We are facing a mobile data tsunami
Hill did not comment directly on the woes over at AT&T, though he agreed that iPhone users consume a lot more data than other smartphone users. On this front, Gene Munster, a senior securities analyst with Piper Jaffray, summed it up most accurately when he told the NYT that, "They [iPhone users] don't even realize how much data they're using."
The main issue, according to Hill, is that the current mobile infrastructure as a whole is not well architected for data. In his own words, the "architecture is wrong." Sketching out the key components in the core network of a mobile phone provider, he highlighted the many pieces of hardware necessary to process and route data coming in from the cell towers.
In a nutshell, the current solution doesn't scale well, forcing mobile phone providers to purchase a lot of expensive equipment just to sustain the "tsunami" of data flowing through the core network. Unfortunately, this exponential deluge of data means that it is a losing proposition for mobile providers that typically sell data packages at a flat monthly rate and is a somber reality faced by mobile operators around the world.
On a side note, Stoke is currently working what the company calls a "Mobile Data Offload" appliance to address this issue. The idea is to create a transparent device that will allow data to circumvent the core network. Hill says that the prototype will be out by mid-October, and the solution should be launched in the first quarter of 2010.Individual data usage
In the course of our meeting, I gleaned an interesting nugget of information pertaining to individual data usage. According to Hill, the average amount of mobile data transferred by a mobile broadband user — laptops via a USB dongle — clocks in at a staggering 3.5GB per month. Obviously, data usage on smartphones will be much lower, though it is not hard to imagine why the iPhone with its superior browser is easily the champion guzzler of data bandwidth.
Anyway, 3.5GB of data per month is a lot when you think about it and definitely worth taking into consideration if you intend to travel with a data dongle or plan to tether your smartphone but not on an unlimited plan.So what does these all mean for IT professionals?
Much as there is a definite convergence in terms of mobile data and the traditional computer network, it is important to recognize the current architecture bottlenecks inherent to mobile data networks.
As such, it is important not to treat the mobile network as we would with, say, a dedicated leased line. As mobile providers re-architect their core networks to be more data-centric and efficient, the situation will probably evolve and improve over time. For now, though, it might make sense to acquire smartphones with Wi-Fi as a fall back for data-intensive or time-sensitive applications.
What kind of data usage do you normally use your smartphones for? Have you ever been stymied by poor mobile data performance?