Technology, Thoughts & Trinkets

Touring the digital through type

The Consumable Mobile Experience

We are rapidly shifting towards a ubiquitous networked world, one that promises to accelerate our access to information and each other, but this network requires a few key elements. Bandwidth must be plentiful, mobile devices that can engage with this world must be widely deployed, and some kind of normative-regulatory framework that encourages creation and consumption must be in place. As it stands, backhaul bandwidth is plentiful, though front-line cellular towers in American and (possibly) Canada are largely unable to accommodate the growing ubiquity of smart devices. In addition to this challenge, we operate in a world where the normative-regulatory framework for the mobile world is threatened by regulatory capture that encourages limited consumption that maximizes revenues while simultaneously discouraging rich, mobile, creative actions. Without a shift to fact-based policy decisions and pricing systems North America is threatened to become the new tech ghetto of the mobile world: rich in talent and ability to innovate, but poor in the actual infrastructure to locally enjoy those innovations.

At the Canadian Telecom Summit this year, mobile operators such as TELUS, Wind Mobile, and Rogers Communications were all quick to pounce on the problems facing AT&T in the US. AT&T regularly suffers voice and data outages for its highest-revenue customers: those who own and use smart phones that are built on the Android, WebOS (i.e. Palm Pre and Pixi), and iOS. Each of these Canadian mobile companies used AT&T’s weaknesses to hammer home that unlimited bandwidth cannot be offered along mobile networks, and suggested that AT&T’s shift from unlimited to limited data plans are indicative of the backhaul and/or spectrum problems caused by smart devices. While I do not want to entirely contest the claim that there are challenges managing exponential increases in mobile data growth, I do want to suggest that technical analysis rather than rhetorical ‘obviousness’ should be applied to understand the similarities and differences between Canadian telcos/cablecos and AT&T.

Ars Technica recently provided a delightfully transparent, technically savvy, piece concerning the problems O2 in the UK encountered with the iPhone, and extrapolates that insight to the problems before AT&T. Central to the problems on both networks is how the iPhone (and related smart devices) actually conserves power. Specifically:

Most devices that use data do so in short bursts—a couple e-mails here, a tweet there, downloading a voicemail message, etc. Normally, devices that access the data network use an idling state that maintains the open data channel between the device and the network. However, to squeeze even more battery life from the iPhone, Apple configured the radio to simply drop the data connection as soon as any requested data is received. When the iPhone needs more data, it has to set up a new data connection.

While this has the benefit of saving on battery power it can have the effect of overloading the signalling channels that the devices use to communicate with cellular towers. While network equipment can dynamically shift spectrum to signalling channels as needed to mitigate the problem, the network nodes may not be able to set up data sessions or get valid network addresses from overloaded DHCP servers. Ars’ source at O2 states that “data capacity is rarely the problem—nodes themselves can usually handle much more data than is flowing through them. However, the networks need to be configured to handle a growing number of devices connecting and disconnecting at a much higher rate than they’ve been accustomed to.”

To be a bit more technical, we can turn to what a person working in the telecom industry stated in the comments associated with the Ars article:

The iPhone breaks 3GPP compatibility in the way it operates (going from Cell-DCH/Cell-FACH directly to IDLE cutting the connection to the network). Keepalives in the applications then cause the phone to open a new connection to the network after only a few seconds. In a normally operating phone, it would just continue using the existing connection since the connection is not cut in between. So the main blame to be placed lies in the iPhone. It doesn’t follow spec. Imagine you making phone calls and hanging up every time, your talk-buddy stops talking. Then dialing again to respond. That’s effectively what’s going on. Other phones (except the new kids on the block following Apple’s lead) don’t hang up in between. The talk is that signalling increases are in the thousandfold range (that’s hearsay though).

These are both technical responses and considerations of the problems facing the AT&T and other global networks: in neither the Ars article where they reached out to technical folk, nor in the case of the telecom worker speaking out to Ars, nor in a similar piece by NSN do we see a statement like: mobile operators need spectrum because otherwise they will end up with the AT&T situation. Spectrum, while helpful in somewhat alleviating the problems associated with the non-spec way that Apple and other smart device OS developers have set up power savings, is not an actual solution. What is required is a combination of updated towers/network equipment and terminals (i.e. phones) that actually comply with international specifications.

What can we take away from this point? First, we need to hear from the technical experts actually responsible from configuring the network equipment and analyzing failures of service. In listening to these individuals some light might be shed on the marketing platforms and rhetoric which is used to obfuscate the actual situations that mobile operators find themselves in. Second, we learn that backhaul bandwidth isn’t really a problem; bandwidth is, in fact, reasonably plentiful regardless of the substantial hike in mobile data traffic.

Unfortunately, without some light being shed into the debate on bandwidth caps and spectrum, we will sleepwalk into a world where the regulatory-normative conditions for participating in the mobile environment will favour limited, and specific, forms of content consumption. Creation will be strictly limited because of tiny upload caps and incredibly punitive overage charges that are drastically out of line with the realities of data transport costs.

In the case of AT&T, again, there is a variable cost of data that indicates 1GB of mobile data either “costs” $75/month (based on the DataPlus plan, where 200MB of data costs $15) or $12.50 – $10/GB (based on the DataPro plan, where the initial 2GB of data works out to $12.50/GB, and each additional GB of data incurs a $10 surcharge). Augmented reality, the integration of high-definition video in smart phones, application stores, streaming video and music, and other high-data services that include both downloading and uploading data are soon to be prohibited by a socially accepted rhetoric of ‘bandwidth hogs’, a rhetoric obfuscating technical realities and threatening to stifle real innovation in the mobile environment. Moreover, with the insistence that wireless can only be a side-channel to a ‘full’ broadband experience (read: wired broadband) telecommunications operators are able to ‘double dip’ and prevent people from cutting the cord and going entirely wireless. Worryingly, the adoption of the rhetoric that is immersed in the pseudo-truths of CEOs and marketers appears to have captured regulators who are not protesting the absurd costs of data traffic that is being imposed on relatively ignorant customers.

One might be inclined to say that mobile operators should be able to charge whatever the market will bear, but this ignores regulators’ unwillingness to let the market bear incredibly high prices for food or fuel (housing is another matter entirely…). While wireless isn’t yet at the same level of importance as food or fuel, the movement towards the Internet of Things, where devices will be dependent of wireless transmissions to provide basic services while consuming (seemingly) precious amounts of bandwidth, threatens to bankrupt a homestead. Broadband is rapidly becoming a utility in Canada and the United States, insofar as essential services are provided over the Internet and there is a rapid march towards getting customers and citizens to adopt a wireless existence. Requiring citizens move towards online forms of engagement, marketing services based on their wireless connectivity, and then turning around and not protecting the customer from price gouging is simply wrong, and the government should examine the actual competitiveness in mobile ecosystems as well as regulate the permissible rate of return on mobile data provision.

As the economics of mobiles are presently oriented the adoption of wireless by either Canada’s or the United States’ citizenry is a poison apple: wireless might be a delightful experience for the first 30 days or so but the 31st day (bill day!) threatens to cripple smart phone users with punitive and seemingly non-uniform data pricing (e.g. 1GB on AT&T costing either $75 or $12.50). Imagine receiving a bill for 40GB of mobile data use…that came in at $405! A wireline connection provided by most Canadian ISPs will grant at least 20GB/month more than that at a cost of $35-45. While slightly higher costs may be incurred for wireless over wireline broadband, the present scale of difference is disproportionate to the actual costs (and challenges) that are actually faced in either of these domains, and especially so in the wireless space.

Unless an accessible and technical outline of why mobile data has to cost an order of magnitude more money to deliver a 33% smaller amount of data than wireline broadband transmission, customers and regulators alike should demand a forced correction of present ISP pricing practices and begin holding mobile vendors to the fire, just as we do oil and gas companies whenever absurd charges for gasoline and oil are witnessed in the marketplace. Failure to be proactive in this space, failure to protect the consumer and provide an ecosystem for application developers to create the next killer mobile app, and failure to hold ISPs accountable for their social responsibilities to provide fast and cost-accessible mobile broadband, threatens to transform North America into an innovation ghetto, where innovation might emerge but is too expensive to actually be implemented and enjoyed on the continent from which it is born.


  1. (*Note, this comment is more of a rant than a true contribution to the discussion)

    Solution! Nationalize the infrastructure! I say that only in half-jest.

    I guess I shouldn’t be surprised anymore that companies willingly break accepted use of specifications for their own purposes (Apple=EVIL I tells ya! *grin*).

    But my god, keeping up with unforeseen implementations of technology that ruin your whole day just grinds you under. My sympathies are with the carrier engineers on this.

    It is weird how information technology is just a little bit different. If someone overloaded a bridge beyond known specifications, and the bridge collapsed the fault would be with the over-loader, not the designer (assuming a competently designed and built bridge). Use a 1-ton pickup to pull a transport trailer is obviously stupid. But with technology, similar usages dictate a need to “innovate” and “keep up with demand”.

    It is tales like this that constantly make me glad I work on a private company network. I can just say no, or alternatively, YOU get management to approve the budget for the redesign. You want it that bad, you be the messenger!

  2. Interesting article. Something that passed by desk recently regarding mobile vs. dsl per GB costs is the following paper from Ericsson:

Leave a Reply

Your email address will not be published.