Can femtocells overcome the barriers of physics?

In arguing the femtocell's case before the LTE World Summit last month, Simon Saunders, chairman of the Femto Forum, invoked the name of one of telecommunication's great theorists: Claude Shannon. For those of you who don't know, Shannon was one of the great minds of Bell Labs back in 1940s and ‘50s, where he developed what we know today as information theory. His theory governs how data is transmitted between two points, and one of its central tenets is that there is a fundamental limit to the amount of error-free data that can be transmitted over any given channel. That barrier is known as Shannon's limit.

Saunders wasn't offering up any academic treatises on physics in his presentation in Berlin last month. He was trying to spell out the numerous advantages of femtocells to carriers and end users alike: increased coverage, offloading traffic from the macro network, even the possibility of femtocell-based applications and services. Though a minor point in Saunders' argument, it was a key one: The industry is fast approaching Shannon's limit. With the maturity of orthogonal frequency division multiplexing access technologies such as WiMAX and long-term evolution (LTE), we'll be bumping right up against that fundamental barrier, meaning any attempts we make to shove more bits into a single hertz of spectrum will be canceled out by noise.

Every few years or so carriers are doubling the capacity of their networks through the implementation of some new technology. AT&T is doing it today with upgrades to its high-speed packet access network, and Verizon Wireless will far more than double overall capacity next with its initial rollout of LTE. But at some point those capacity gains will stop. In order to grow network capacity, carriers will have to devote to 4G more spectrum, which itself is a limited commodity. Ultimately the only way to grow capacity will be through spectrum re-use, i.e. shrinking cells. And here's where Saunders believes the femtocell will shine. After all, the femto is the ultimate example of the shrunken cell — a personal base station for the home or office.

It's easy to jump to the conclusion that Saunders is making dire academic predictions in order to sell the product for which he's the industry's primary advocate. But few scientists and engineers in this industry would disagree with the premise, though they may question his proposed solution. Håkan Eriksson, chief technology officer for Ericsson, has long talked about the limits of future radio technology and has even pointed out that the incremental gains the industry can still achieve in capacity are offset by losses in cell range. Smaller cells are inevitable in the networks of the future, Eriksson believes.

Femtocells may not be the answer, though, and if they are, they'll only be part of the answer. Millions of cheap femtocells could replicate the coverage of thousands of base stations in a metro network and add untold amounts of cheap capacity to boot, but they would create a networking planning and management nightmare. Saunders definitely is onto something, though. The industry began by creating the largest possible cells and has gradually been shrinking them. The Femto Forum is starting with the smallest possible cell. At some point they're going to meet somewhere in the middle.

E-mail me at kfitchard@telephonyonline.com.