Adding a carrier-grade femtocell can add capacity to the network. Of course, adding another sectorized macrocell on a new tower can add even more capacity. Which method is the most economical way to add raw LTE data capacity to the mobile network?
In our recent interviews for the Mobile Experts Small Cell forecast, 26 different mobile operators (three-fourths of the total) said that they will be adding LTE to an existing network. This means that they already have a grid of towers and rooftop sites for 2G and 3G services, so they plan to begin with macrocell coverage, originating from these sites. It's clear to many of them that they will have coverage holes and that capacity may be filled quickly. So many operators are already looking ahead to the next step: How to add capacity when the existing towers are full?
Not surprisingly, almost every item is more expensive for the macrocell case than for a femtocell. The tower and the high level of power consumption for a macrocell require much more infrastructure, and the height of the tower means that the mobile operator must carefully plan the RF propagation for each tower. High throughput at a macrocell site normally dictates expensive backhaul: fiber or line-of-sight microwave links.
On the other hand, small cells can be inexpensive. The high level of integration at the silicon level has multiple impacts:
- Lower cost for the equipment itself
- Less heat, eliminating heatsinks and air conditioners
- Smaller equipment, requiring no footprint on the ground
- Pole-mounted or wall-mounted equipment requires little cabling and inexpensive antenna
- Cheaper core network support results from the femto/gateway architecture and the use of local Internet breakout
In addition, the use of Non-Line-of-Sight backhaul in small cells could have a significant impact on backhaul cost. The initial installation will be quick and easy. The licensing can be done at a nationwide level instead of "one site at a time." And the licensing itself could be cheaper than LOS microwave systems.
Looking at the capacity of macrocells vs. small cell systems can be surprising. While early femtocell systems were designed to be low-capacity "hole filler" products, more recent small cell designs include baseband processors with up to 48 processing cores. There's a lot of capacity built into 2013 carrier-grade small cells! In fact, because the radio conditions can be better for small cells than for macrocells, we expect that a well-placed small cell will handle more capacity than a single macro sector. For 20 MHz LTE bandwidth, a small cell may achieve 2.0 bps/Hz, while the macrocell would achieve 1.7 bps/Hz under typical radio conditions.
Results for individual sites
For a network that already has widespread LTE coverage, adding another incremental chunk of capacity is much cheaper for small cells than for additional macrocells. Each Mbps of additional capacity would cost $3,700 if an additional tower was required, but only $1,200 for small cells (total cost of ownership over 8 years). The conclusion: If traffic is centered around hotzones where small cells can be effective, small cells are a winning solution.
The overall conclusion of our study was that despite the very inexpensive incremental cost of small cells, in general a new tower is a better choice whenever eight or more small cells would be required to match the same user experience. Any real-world upgrades that anticipate strong data growth should incorporate both macrocells and small cells to achieve the best performance and lowest cost.
Joe Madden is principal analyst at Mobile Experts LLC. He is a Silicon Valley veteran, with 23 years in mobile communications, navigating through IPOs, LBOs, divestitures, acquisitions and mergers. He leads the analysts at Mobile Experts, focusing on nitty-gritty analysis of mobile communications and semiconductor markets. Madden graduated, cum laude, from UCLA in 1989.