The Long-Term View on Small Cells
February 12, 2014
The evolution of wireless networks to a HetNet architecture is inevitable, but the question for industry players and investors is what form will it take. This is because there is no single approach to small cells, but rather there are multiple ways to densify the network. Making misplaced bets is common in the technology space, mainly because of the myriad of cause and effects that take place where technology is only a factor among many other factors, and at times a minor one in comparison.
When looking at the small cells and hetnet space, there is a trend that we can follow by a having at sober look at snippets of information that we can gather despite much noise. One of the main emerging trends is greater coordination between the macro and small cells. Coordination is in itself a broad and general concept with many different techniques and implementation options. This is a subject that the 3GPP has picked up on in defining LTE release 12 features which are centered on enabling mass-scale hetnet deployments. In a major architectural change, the user plane and control plane will be separated. In layman’s terms, a macro cell will control the small cell and traffic can go to either small cell or macro cell.
There are implications to this architectural change that will reverberate backwards to the core network. The requirements on backhaul will have to become tighter: faster, more responsive and more predictable backhaul performance is required (in other words, higher capacity, lower latency and lower jitter). Moreover, fronthaul will be set to grow as coordination is simplified through aggregation of baseband resources along the concepts of Cloud RAN. This is specifically applicable to outdoor deployment scenarios as it provides superior performance over distributed techniques (the business case in some scenarios is doubtful – but that’s a whole topic in itself!)
The other important piece is that of spectrum for the small cell layer. The greatest benefit comes when the small cells are deployed in their own spectrum layer which minimizes the interference scenario at the cost of additional spectrum resources. The evolutionary path today is that 3 GHz bands (3.3 GHz in Asia, 3.5 GHz in Japan and in Europe with extension to 3.8 GHz in certain markets) will be the small cell band. Trials in Japan have been in progress in this band. For the United States, the option is to leverage dynamic access spectrum in 3.5 GHz – the FCC has been very active on this front with firm intent to free as much as 100 MHz for tiered spectrum use. Another alternative will be to implement LTE in 5 GHz band which was proposed by Qualcomm and is likely to be accepted for Release 13 activities once initial resistance to this plan is reduced. In short, the bands between 3 and 6 GHz will be those for carrier managed small cells much as Wi-Fi today dominates 2.4 GHz and 5 GHz bands.
The evolution of hetnets can take on different path. In one path, we have carrier managed small cells where scalability is bounded by being able to cost-effectively manage accessibility to the network and quality of service. In another path, we have Wi-Fi which compromises on performance in comparison to licensed-band small cells in favor of flexibility borne out by its inherent architecture and spectrum regime. In that sense, we can expect the search for the golden mean to continue for the foreseeable future.