Which spectrum will have the most cellular base stations installed between now and 2020?
The global patchwork of cellular networks is changing radically, as operators embrace significant changes in thinking about how best to structure their base station operations, especially as LTE-Advanced emerges.
Most operators will build out starting with a coverage-oriented macro layer, which is vital for VoLTE. A few frontrunners have already moved on to the ‘densification’ phase, involving a high spend on small cells, WiFi and other technologies in separate spectrum bands, licensed or unlicensed.
There will be a rising investment in software to allocate and share radio resources efficiently, including optimization and offload tools. The more intelligent and agile networks become, the further operators are moving towards fully software-defined networking (SDN). One aspect of this is virtualization, in which multiple cell sites share the resources of a single baseband unit, either in a nearby ‘hotel’ or remotely in the cloud. The increasing decomposition of the base station, which began with remote radio heads and continued with integrated radio/antenna units, has its logical extension in Cloud-RAN or virtualized RAN (vRAN), though this is in its early stages of commercial deployment.
But by 2018, about 40% of the world’s cellcos will have vRAN deployments in relevant areas, but three-quarters of these will be using a hybrid architecture which combines Cloud-RAN and traditional techniques in order to reduce reliance on fiber and to preserve investment in older technologies.
Operators are evolving the vRAN idea further too, so by 2018 there will be ‘Super vRAN architectures’. These typically add further scalability and cost efficiency by extending cell sites’ reach using distributed antennas; and by adopting highly flexible architectures in which cells dynamically come on and off, or change size, in response to demand at any one time.
Initially at least, the main attraction of this vRAN architecture is the promise of a very low-cost cell site, which could be squeezed down to $100 by the end of the decade.