Mobile Entertainment

A fundamental paradox exists in the mobile video marketplace. Call it the Law of Spectrum Scarcity. At first blush, everything mobile video is coming up roses. Viewers are on the rise, revenues are up (198 per cent year over year according to Telephia), and projections are through the roof.

Clearly, operators see mobile video as a long term, money making machine: more services, more viewers, more dollars. The problem is the traditional equation doesn’t work: to increase the number of viewers, you have to either increase the system bandwidth or decrease the video quality per user.

Adding bandwidth is slow and costly. It requires the building of new cells and backhaul facilities, which affect the bottom-line. But decreasing video quality obviously threatens user adoption.

So how can this be addressed? One way is to add an overlay network. MediaFLO, DVB-H, S/T-DMB, and ISDB-T are some of the standards being deployed to broadcast television to mobiles on an overlay frequency. While this method requires a separate radio in the handset, the high quality and predictability of content makes this method attractive for users who like the idea of watching live TV and are less concerned about downloading content. The drawback, however, is the ‘scarcity of content’ the broadcast technologies introduce. In other words, you only get to watch what’s on – you can’t choose your own content or dictate when you watch it.

The fact is, mobile video has evolved based on an internet model, not a TV model. Users expect to access video on demand (VOD) and user generated video. And if it’s true that VOD is the real demand driver, but also the dreaded spectrum-eater, then mobile video may not be the panacea operators have been looking for.

Fortunately, there is another option: make unicast work in a way that doesn’t threaten the spectrum. After all, just because the model is based on the internet, doesn’t mean the technology should be. Take streaming video. Here, a mobile handset is simply seen as another internet device – as if it were tethered to the tower with an invisible high speed data wire.

In this instance, video streams are pre-encoded with bitrates based on specific handset capabilities and the type of network they’re on. This is the equivalent of adapting internet video feeds based on whether the user is on dial-up, DSL, or an office LAN.

Yet wireless networks are completely different beasts, because the bandwidth constantly fluctuates based on changes to user location, interference, subscriber density, distance, etc. Using the analogy above, it would be like trying to send video to a target user that is on an office LAN one second, then drops to dial-up and back to DSL.

So these internet-style solutions are unfortunately taking a lowest-common-denominator approach to mobile. The resulting video can be ‘just good enough’, but such an approach is simply lowering the denominator and unnecessarily limiting the quality of the viewing experience.

What is required is a new approach, which expands the reach of the existing spectrum, while improving video quality. Not easy! The key is to look at several variables within a video stream –including initial latency, frame quality, stalling, re-buffering, and audio quality – to determine if there’s a way to manage them in such a way as to positively affect the user’s perception of quality, while only using the available spectrum. In other words, change the equation.