The higher-than-expected prices in AWS-3 spectrum auction are an indicator of the demand that operators foresee for LTE capacity. But spectrum is inherently scarce and, as we are seeing, increasingly expensive. Furthermore, owning spectrum is not the same as delivering it to subscribers, who are more often than not indoors.

In the best circumstances cellular signals don't penetrate building walls well, and the high frequency AWS-3 bands up for auction are especially poor at this.

Small cells bring capacity to subscribers, often re-using existing spectrum. This is particularly critical in dense urban environments where the spectrum capacity crunch is worst. Unfortunately, outdoor small cells mounted on poles or lampposts are expensive in many aspects: rights negotiations, backhaul, power and ongoing maintenance. Outdoor small cells also can create interference and frequent hand-offs with the macro.

A better alternative is indoor enterprise small cells. They can be deployed in offices, public buildings and multi-dwelling units (MDUs) that are common in cities. They are inexpensive to deploy and maintain since most enterprises and MDUs will welcome better wireless coverage and will share power and backhaul for such improvements.

Consider that there is a bid of $2.76 billion for one slice of New York City area spectrum. For the same investment, an operator could buy 1.35 million enterprise LTE small cells at an average fully installed cost of $2,000 each. That's one LTE small cell for every 18 people in this market, or for every six subscribers of an operator with 33% market share. Moreover, small cells typically support 20MHz, double the bandwidth that the operator is getting in the auction. Most important, enterprise small cells deliver the capacity where the users and usage are: indoors.

For a mobile operator, one million small cells in a single metro area is a daunting prospect. Operators have expressed concerns to us about the potential for interference with the macro network. Some have also told us they have difficulty managing installations of more than four or five standalone small cells within a single enterprise, due to border interference between the small cells. However, new approaches are emerging that promise to make large-scale small cell deployments a more attractive alternative.

Cloud Radio Access Networks (C-RAN) has been an active topic in the macro cell community since 2011. Now the concept is being applied to small cells. Last summer the Small Cell Forum announced a work stream to "understand how small cells fit into the Cloud-RAN story," focusing on architectures that centralize baseband processing across multiple access points. This allows access points deployed across a large building to form a single cell, eliminating inter-cell border interference and handovers, and it has significant implications for small cell performance and economics.

We've seen firsthand that coordinating the scheduling and Layer 1 functions across access points can provide up to 10x faster data rates and a 5x decrease in jitter (to ensure VoLTE quality) compared to traditional standalone small cells, while also improving battery life. From a macro interference perspective, this approach vastly simplifies SON coordination by reducing the number of small cells that an individual macro "sees" from tens or hundreds to just one. And from an installation perspective, because a C-RAN system acts as a single cell, the complexities of handover boundaries and inter-cell interference go away.

Unlike indoor distributed antenna systems (DAS), these emerging C-RAN Small Cells can be deployed in much the same way as managed WiFi systems are, using standard Ethernet LANs for in-building distribution. This drastically reduces the material and labor cost of installation. The high cost of DAS equipment and installation explains in part its low penetration of only 1.3% of North American office buildings today, according to ABI Research.

Current-generation small cell management systems also reduce complexity and increase reliability of small cell installations. These systems import macro cell site parameters and dynamically configure the small cells to minimize interference and improve handovers with the macro network. Instances of nearly 1,000 small cells operating in the footprint of a single macro cell, without impact on either macro or small cell performance, are a reality today.

LTE continues to evolve, and thoughtfully designed C-RAN small cells will support LTE-A capabilities such as Carrier Aggregation, Coordinated Multipoint (CoMP) and distributed higher-order MIMO, all techniques that are critical for delivering high speed data services. Small cells are also capable of determining user location within a building, a necessity for public safety and location-based services.

The incessant growth in demand for mobile data services makes new spectrum purchases inevitable, but small cells with advanced C-RAN capabilities create a new opportunity to meet that demand more economically and also deliver the best possible user experience to subscribers.

— Michael McFarland, Senior Director of Product Management and Marketing, Airvana Inc.