With the rapid takeup of smartphone applications and cloud computing, you’d think there wasn’t a place left for femtocells to differentiate. Here we look at the different places where the intelligence for femtocell applications can reside, and trade-off the benefits and disadvantages.

Where can you find the intelligence?

IT architects have been arguing about the best structure for applications for decades. Thick client, thin client, cloud computing, intelligent networks…. these all have their own model with strengths and weaknesses.

In the end, it seems to me that the intelligence (or logic) for any application can reside in one (or more) of several places:

1. On the device itself. There are many standalone applications for iPhones that don’t need to communicate with anything else – Games such as Suduko or Chess might be a good example.
2. In the cloud. Accessed using a browser or simple client, often large data banks can be filtered and accessed for specific content – streaming audio such as spotify or last.fm, google maps might be good examples of this.
3. In the network. Some information is only known within the operators network, and logic can be applied even when you are disconnected. Call handling and voicemail might be good examples of this.
4. Inside the femtocells: Although perhaps considered as an extension of the network, femtocells can offer presence information including whether anyone is at home or not – combined from multiple users.

It could be argued that the mobile industry has had the opportunity to offer a wider range of service logic but moved too slowly to capitalise on it. One example would be location, where some years ago many innovative location services were being suggested. Instead, device manufacturers have integrated GPS which when combined with Google Maps or other cloud services provide a compelling and growing choice of features - completely bypassing the network operator. It’s now unlikely that location information can be sold by operators cost effectively and has instead become an extra burden for emergency call regulations.

Some of the best services use a combination of two or three of the above - commonly a mixture of device and cloud. Sharing and synchronising your contacts and email between devices, ebook readers and many other applications benefit from this approach. Up to now, few have direct interworking with the mobile network operator’s services apart from the standard voice calling, text messages and voicemail.

Femtocell Services can’t win by being standalone

The industry has racked it’s brains to dream up and prototype innovative new femtocell applications. Some are commercially deployed, such as NTT DoCoMo’s alert service which emails or texts parents when their children get home from school. Most that I’ve see so far are pretty much standalone and don’t involve a device application or cloud server except for configuration purposes.

While this means that they will work with any phone, it limits the potential.

Opening it up to the new breed of app developer

I believe that opening up the network API to 3^rd party developers is essential to enable the widest range of innovative applications. If this was presented simply as an extension to existing iPhone, Android and other development environments, it would drastically reduce the barriers for “femto enabled” applications – development cost, time to market and accessibility would all be greatly improved.

The alternative of individual operators hosting their own unique set of femtocell applications seems to me to be doomed to failure – it wouldn’t attract the critical mass of 3^rd party developers to innovate, making it likely to be expensive, slow to market and risk being bypassed by alternatives.

It’s the end-to-end customer proposition that will win

Apple have set the standard for customer experience and ease of use for mobile applications. Their iTunes Appstore provides a simple and very convenient way to buy and install new applications. Developers, although they have to go through an approval process, get paid a reasonable revenue share and can deal with a single primary customer (at least for billing and cashflow purposes).

This compares favourably with a situation requiring sales relationships with many different network operators and a huge variety of different phone models. By reducing the complexity and increasing the addressable market, the iPhone Appstore has set the standard. A similar effect has been achieved with Google Android and RIM Blackberry.

The femtocell industry needs to specify a small set of APIs and find a way to make these available for and accepted by developers who use today’s leading application development tools.

Comments (0)