Avren's annual basestation conference covers a much wider range of topics than just Femtocells, with vendors touting everything from complex shared tower sites through to the latest LTE small cells. Strategists from various operators shared their thoughts on how networks will evolve with the relatively new terms HetNets and small cells gaining ground.
I won't detail every last presentation – you'll have to attend the full event for that – nor will I be able to share much of the useful side conversations and networking that accompany this type of event. However I have written up a few of the key points and taken some video snippets with several participants.
The sun was shining, the hall was packed and the 3 day event has a pretty full agenda. There was a good mix of operators versus vendors, with a few unobtrusive exhibitor stands. The event concentrated on open sharing of views with a strong technical element.
Here's a short selection of a few topics covered during this first day:
What's the cost of a Gigabyte of mobile data?
David Haszeldine, Network Economics Manager from Deutsch Telecom, did his best to explain and justify an answer to this frequently asked question. Without giving an exact figure, it seemed that operators could easily be making a loss if they weren't careful with their data plan pricing (or if their assumptions about average consumption were optimistic).
He made were several non-obvious assertions:
- 50% of network traffic is carried by 10% of the cellsites. The demand curve was quite extreme, with two conclusions:
- There are many non-profitable cellsites, typically in rural areas to achieve coverage. These aren't worth expanding and won't be upgraded to IP/Ethernet backhaul anytime soon. Site sharing is most worthwhile for these and actively pursued by all operators.
- The ratio of 10% busiest cellsites/50% traffic was true for each class, i.e. 10% of rural cellsites are also carrying 50% of the rural traffic.
- Data traffic is no longer growing exponentially; growth rates are starting to roll off (ie still growing, just not as fast as before). This may change as new iPads, tablets and other gadgets appear in the future.
For me I took this to mean there are a small number of high traffic areas where small cells, including metro-femtocells, are going to be targeted to satisfy that high and growing capacity. In rural areas, David suggested that low cost rural femtocells might be justified in a remote hotspots – much more cost effective than simply building out another full size cell tower.
HetNets and Small Cells are the new buzzwords
Huawei expounded their view of the utopian network of the future – using the term HetNet (Heterogeneous Network). This is a combined mix of macro/pico/femto and relay cellsites using 2G/3G and 4G in a co-ordinated system. They have developed a feature called SingleSON (single self optimising network), which extends their SON capability developed for LTE to operate with 3G and GSM too. This automatically reconfigures and tunes the network to match demand and the environment. From their viewpoint, cellsites must be simple to install "plug and play" while operating very efficiently.
It seems that all of the large basestation vendors are now saying that femtocells (whether they call them small cells, picocells or something else) play a part in meeting the traffic capacity demands of the future. These small cells also deliver high data rates achievable when operating close to a basestation and give a much better end user experience. One operator told me he thought that it was the low latency of LTE that would be most impressive to consumers rather than just the high peak data rates because it would give a very fast response to any query, such as calling up a webpage.
Who's going to pay for LTE?
At a panel session, Telefonica pointed out that network operators have a finite amount of money to spend – so if they are to invest heavily in LTE, then either they have to stop spending on something else or charge more.
This may be a pretty obvious statement, but worth repeating – the (not insignificant) cost of rolling out LTE has to be found somewhere. There have been some significant savings in recent years though - one of the speakers (from RadioDesign) explained how site sharing has dramatically reduced costs by halving the number of cell towers, with their clever RF filters being used to allow fewer shared antennae.
There are clearly some difficult strategic decisions to be made by network operators as to which types of equipment and technologies they continue to invest in as their choices continue to expand.