T-Mobile & femtocells

T-Mobile announced this week that it’s running femtocell trials in the UK, Germany and the Netherlands. According to reports, Emin Gurdenli (Technology Director at T-Mobile UK) said “A proper food chain has to be established before this thing rolls out in large numbers. But I’m personally optimistic.” He also said that proprietary systems are “not viable”.

I was lucky enough to hear more from Mr. Gurdenli at the Marcus Evans femtocell conference in London this week. Although he did say that standardisation is key for femtocells to take off in a big way, he also said that proprietary, non-standard solutions are “OK for a year”. This is in line with Carl Weinschenk’s comments last week.

As befits a technologist, Mr. Gurdenli presented the key operator benefits of femtocells largely in terms of increased coverage & capacity, and offloading the macro network. He showed the chart above (originally due to Gabiel Brown), and commented, “this is where I lose sleep”.

Other take-aways from the presentation include:

• Customers need the highest data rates in-building, but the macro network delivers the opposite.
• As far as the technical challenges of deploying femtocells are concerned (interference, plug-and-play provisioning, location etc. etc.), Mr. Gurdenli expressed great confidence that there is “nothing insurmountable”.
• T-Mobile is seeing an explosion in data use. As data volumes grow, backhaul costs account for 75 percent of the expansion costs needed to deal with that growth. Femtocells use the customer’s existing DSL (or cable) connection for backhaul, of course, so that solves the problem in the home. However, elsewhere last week Mr. Gurdenli commented that T-Mobile also plans to migrate its macro network backhaul traffic away from expensive E1 connections to DSL. (IP backhaul over DSL lines is already used widely with picocells, but it’s relatively new for macro base stations.)

When asked to comment on whether T-Mobile International (with its femtocell trials) is going down a different line from T-Mobile US (which appears very committed to its HotSpot@Home dual mode UMA service), Mr. Gurdenli commented that HotSpot@Home is aimed at solving coverage problems, whereas he sees femtocells as a platform for service innovation.

In fact, this was a key theme of the conference. There was a lot of discussion about the idea of femtocells enabling new services that are only available when the mobile phone is in the “femtozone” at home. More on this in a future post.