In the last quarter of 2010, global shipments of smartphones have surpassed desktop computers and notebooks for the first time ever.
No doubt, this was an indication; an inflection point in technology and communication industries that mobile devices were fast emerging as the primary end-user access points for information. In India, more people access the Web through mobile phones than in any other way.
Mobile games and applications in India are expected to become a Rs. 2,700-crore market by 2016. It is also estimated that two-thirds of the world’s mobile traffic will be through videos by 2017.
With growing popularity for videos, trailers, movies, serials, and user generated videos, the story will not be very different in India.
Such trends come with a major challenge for telcos: to ensure high user ‘quality of experience’ for content consumption, while keeping their capital and operational costs low.
Interestingly, networks often observe significant variation in utilisation levels, mainly triggered by diurnal patterns of human activity. Such variations cause low utilisation at certain times of day and overload during others. Overloaded network conditions are bad for the user quality of experience, since they can lead to video stalls, jittery videos, or just low quality videos. Solutions that can help telcos smoothen network traffic could improve the “yield” of the network provisioning that is already done, before upgrading it. The Indian wireless operators, in 2010, spent Rs. 1,062.19 billion on 3G and BWA spectrum auctions, and are continuing to spend more on network deployments. The operators have to recoup these “sunk” costs from services provided and yield optimisation by moving some traffic to lower utilisation times through price incentives. One promising solution to smoothen traffic will be for telcos to introduce a mechanism for flexible asynchronous delivery of certain kinds of content.
In return for delaying content delivery, telcos can provide price incentives such as lower delivery price per byte. For instance, on each video request, the solution presents the user with options to: (1) stream and watch video at that time by paying the regular price for the transfer, or (2) download and watch the video later at a discounted price with expected delivery time (EDT) options.
In a nutshell, the solution introduces a “video sachet” model, in which users are allowed to choose the delivery price and expected delivery time of different videos at fine granularity. One key observation the sachet model leverages is that user behaviour not only depends on price and expected delivery times, but also on specific content and context of the content.
For instance, a young cricket fan may not be willing to wait even a minute to watch a highlight video, but may be willing to wait for even an hour to receive some other educational content or a Bollywood movie.
At the same time, the solution facilitates a telco — through appropriate price and expected delivery time — choices to control when and at what price the content is delivered to users. This way, it helps telcos reduce operational costs. Such a solution could be applicable to a large number of content types such as video downloads, software updates, multimedia uploads, and so on, which are amenable to asynchronous delivery.
Notably, the telcos do not need to reduce the offered plan cost for all traffic uniformly, as they currently do to beat competition.
The impact of such a solution can be enormous. Imagine delivering educational content based on video for mass markets at low cost. With the right solution such as a fine-granularity asynchronous delivery solution, a telco will be able to extend discounted content transfer for students and teachers to download educational videos on low-cost smartphones or tablet computers. In fact, such a solution may trigger much faster adoption of 3G or 4G technology by larger segments of population.
(The author is Senior Researcher, IBM Research-India)
