The not so holy grail of Mbit/s
Over the course of 2011, I’ve attended several telecoms conferences throughout the EMEA region and worked on numerous projects we’ve conducted for operators. From what I’ve seen and heard over the past year, the whole industry seems to be trumpeting for the essence of focusing on the end-user experience.
And it makes perfect sense to me. I believe that all network optimisation work should be aiming to improve end-user satisfaction – otherwise it’s just worthless engineering with no clear or justified target.
Yet at the same time I can still see a total divergence of promise, delivery and expectation of the single most important technical factor impacting the end-user experience – the bitrates.
Operators (and vendors!) are commonly selling (1) headline speeds while typically only delivering (2) a fraction of what they promise. However, at the same time the end-users expect (3) 10M mobile broadband to correspond with 10M DSL even though we all know that you need to be living in Dreamworld for that to happen.
The fourth point – and perhaps the most important one – is that very few seem to know or care what kind of bitrates are actually required for different services. I’d say that this lack of awareness is exactly what fuels the absurd hunger for continuously increasing connection speeds.
For instance, if we look at the end-user experience of WWW surfing – which is mainly determined by the time spent waiting for a web page to load – we will see that the relation to bitrate is a) highly non-linear and b) not correlating strongly after certain limits:
I know that the throughput is generally considered to be the “the engine’s horse power” and it is certainly important. However, the results above seem to conclude that other measures should be taken as well if we are to approach the true end-user QoS holistically.
In fact, the big buzz around bitrates reminds me of the the frenzy built up regarding megapixels in cameras and megahertz in computer processors a few years ago: we were all constantly striving to get hold of, say, the camera with the highest number of megapixels. Nowadays, we all know that the quality of a photograph or laptop is dependent on many factors, with megapixels or hertz being just one. Provided there are enough pixels or hertz for the equipment to do its job, we now look for other specs that seem to meet our needs.
Actually, the number of megahertz in laptops has decreased recently in order to preserve battery life. This is what came about when CPU manufacturers focused on the end-user experience. Now, I’m not suggesting MNOs to provide lower bitrates, but I would highly recommend that mobile service providers set up a holistic quality and capacity management process instead of making bullish, peak-rate feature rollouts. This process would then drive network rollout and optimisation based on true end-user needs, capturing the most important aspects of WWW, VoIP, Video and other killer applications.
I wish a high quality New Year for all the end-users of Omnitele blog
Manager, Business Development