I spent last week in London at the Small Cells World Summit. If you’re looking for a recap of the main themes, you can check that out here.
If you want a shorter version, the event boiled down to two key themes: old stuff and new stuff. Simple, huh? Much of the vendor and operator messaging focused on issues that the market has been grappling with for a while. The slow maturing of solutions and slowly ramping deployments. Difficulties in actually siting small cells. The evolving definition of what a small cell is. The availability of many, competing small cell backhaul technologies and architectures – with little insight (yet) into which may win out.
It wasn’t all “old news,” however. Two new-ish dynamics came to the fore at the show. We’ll doubtless hear more about them as the year rolls on, if only because they have the potential to impact the market significantly.
- Inside-Out Coverage and Capacity. Traditionally, operators have delivered indoor cellular coverage and capacity from outside the building. Small cells promise to target indoor coverage, but if signals extend outside the home or office, they could deliver broader coverage. Presumably this is part of what BT means when it talks about an “inside-out” strategy for using its new 4G spectrum. It’s exactly what Qualcomm had in mind with its silicon launch, positioned as a solution for “neighborhood cells.” Given signal attenuation through the outside wall, coverage and capacity may be limited. Regardless, turning unplanned small cell deployments into a manageable network will require solid SON assets.
- RAN Virtualization. If you thought the confusion between small cell, DAS and C-RAN architectures was going to dissipate any time soon, you’re going to be disappointed. While not formally announced, the Small Cell Forum used this year’s event to kick off work on RAN virtualization (vRAN); you can check out initial details on the work and its rationale here. Driven by a diverse set of vendors – Alcatel-Lucent, Cavium, Cisco, Intel, and ip.access – the work makes sense as operators evaluate the coverage and capacity options open to them. While it’s unclear how operators might integrate vRAN into their deployments, it is clear that use of CPRI to link virtual (digital) and physical (radio) components simply won’t work for many operators; this generally requires fiber connectivity (or maybe high-capacity wireless fronthaul). Key to the success of the work, then, will be figuring out where virtualization makes sense and how functions can be divided in order to limit the connectivity requirements.