Madden: Cloud RAN or small cells?

Joe Madden mobile experts

     Joe Madden

It's funny to watch the industry hype machine at work. Every new idea that comes along is promoted as the "next big thing." In the case of Cloud RAN, we have seen several vendors promoting their trials and initial deployments, and we have all heard about the savings available in baseband pooling. Cloud RAN will make dense urban networks possible, and putting the baseband processing together will make LTE-Advanced features possible.

At the same time, small cells have also been advertised as the solution to the same high-density traffic problem. Instead of pooling baseband resources, small cells distribute the baseband processing. So which is it?   Are we going to bring the baseband processing together, or spread it out?

Both technologies are promoted as a way to save cost. The Cloud RAN architecture is intended to save operational cost, by locating everything together and allowing a technician to work more efficiently in maintaining baseband resources. Cloud RAN also enables LTE-Advanced features that improve capacity.


It's true that Cloud RAN makes LTE-Advanced easier to implement. Coordinated Multi-Point (CoMP) requires very low latency (less than 1-2 microseconds) to provide a big boost in capacity.   Enhanced Inter-Cell Interference Coordination (eICIC) coordinates blank and almost-blank sub-frames on a real-time basis, so locating the baseband processing in one place greatly reduces the complexity of coordinating changes on the fly. The Serdes interface to a Remote Radio Head is the lowest-latency connection possible, with most of the delay coming from physical propagation instead of IPSec and processing time. As a result, multiple RRH units on a single baseband node offers the highest performance boost from CoMP.

The financial benefit of CoMP comes from the added capacity and higher throughput available at the cell edges. CoMP can improve throughput by as much as 80% at the cell edge, for both uplink and downlink. Using a traditional macro network as a baseline, the overall savings can be very significant in dollar value, because fewer macro eNodeBs are required. Additional savings come from the centralized location of the baseband processing, making maintenance simpler and more efficient.


The Small Cell architecture is also intended to reduce cost per bit, by reducing the overall cost of baseband processing and radio hardware, as well as increasing spectral efficiency and overall throughput. Small Cells can drop the cost per bit by a factor of 4, compared to a macro LTE network. Even with high-cost backhaul such as millimeter-wave links, the cost per bit can drop in half.

So which architecture gives a bigger impact to the guys in Accounting? It looks like the answer to that question is "it depends." The Cloud RAN architecture is favored by operators with access to cheap fiber (China Mobile, NTT DoCoMo, SKT, KT), and, by other operators, only for stadium situations or other localized problems. The Small Cell architecture seems to be popular with most mobile operator CFOs, because the transport cost for RRH fiber outweighs the operational savings of co-locating the baseband processors. As examples, China Mobile and NTT DoCoMo like C-RAN because they own the fiber in their countries. Competitors such as China Unicom and Softbank like small cells because they pay much higher cost for fiber bandwidth.


Will mobile operators discover more value in Cloud RAN when CoMP and eICIC are widely adopted? Maybe.  But the Cloud RAN promoted so far doesn't change the economics of a macro network by a factor of 4, as we expect small cells to do. For wider adoption, Cloud RAN must achieve some significant savings to overcome the penalty of high transport cost. At the same time, CoMP and eICIC will be implemented with small cells, and while the benefit may be smaller, we can expect significant benefits from these features in a HetNet configuration.

In the end, most operators will be pushing ahead with the Small Cell architecture, and may consider Cloud RAN for special situations such as a stadium with very high density. If fiber bandwidth is free, then the cost of transport is no big deal, and Cloud RAN is attractive. For most operators around the world, fiber cost is a significant chunk of their operating budget, so we expect most of the market to move toward Small Cells.  

Joe Madden is Principal Analyst at Mobile Experts LLC. Mobile Experts is a network of market and technology experts that provide market analysis on the mobile infrastructure and mobile handset markets. He provides market forecasts for handset, DAS, small cell, and base station markets, with in-depth research down to the nitty gritty details of frequency bands and power levels. Mr. Madden graduated, cum laude, from UCLA in 1989 and is a Silicon Valley veteran. He has survived IPOs, LBOs, divestitures, acquistions, and mergers during his 24 years in mobile communications.