Manish Singh, CTO of Radisys, shares his industry views on the emerging consensus for small cells to solve the data capacity issue, discusses how different vendors are approaching product development and reveals a surprising aspect of the format and purpose of the first dual-mode 3G/LTE femtocells.

I caught up with Manish Singh, CTO Radisys, who was presenting at Informa’s Broadband Traffic Management Congress in London. The event has continued to grow in popularity with some 500 attendees and attracts support from vendors and operators alike. Radisys provide the core hardware (ACTA) platforms which are used for deep packet inspection and traffic management by many in the industry, as well as the software components which support them. They sell to product vendors rather than the operators directly, but this gives them good/early insight into what the industry is working on.

Manish has been pretty busy this year due to Radisys acquiring Continuous Computing. He has been regularly flying between San Diego (where Continuous Computing were based) and the Radisys office in Oregon. This is in addition to supporting conferences, events and customer meetings around the world. He reports that the acquisition has gone well, both companies had very complementary product sets and the usual integration issues have been worked through.

Is the demand for data still continuing to grow?

At the [Broadband Traffic Management Congress] event, we heard some speakers suggest that perhaps the rapid growth rates for mobile data are peaking. I’m not so sure – we continue to see new smartphone devices, a wider take-up of applications and strong demand for [computing] tablets. These all continue to increase demand and place a strain on networks – a trend that we think is likely to continue. If anything, it is the limited capacity of some networks which is the constraining factor – consumers would use more data where it is available at speeds and capacity that support new types of application as yet undreamt of.

Are femtocells now accepted as the industry solution for mobile data capacity?

I think everyone would agree there is now a clear industry consensus for small cells. While there have been some in the industry who have been sceptical about the femtocell concept, especially in the residential context, we are now seeing a very strong agreement throughout the industry that small cells are the right direction.

There are two approaches to building small cells: downsizing existing macrocell designs or upscaling existing residential/enterprise femtocell products. Depending on your viewpoint, the term small cells can either include femtocells or mean the same thing.

What are the tradeoffs between the two approaches?

Some of the large RAN [Radio Access Network] vendors are taking the first approach, downsizing their existing macro and microcell products. Their mature macrocell designs take advantage of many years of development and testing, and incorporate many advanced technical features. Their challenge is to reduce the functionality, footprint and capacity to achieve the purpose required – namely high capacity at low cost. The problem is that this mature software often includes a lot of functionality not required in small cells, for example handling high speed mobility and handoffs. This can’t always easily be stripped out because it is embodied into many different parts of the codestream.

All of this software takes processing power to run. Every additional MIP required drives the cost up, not just from the silicon it runs on but the power supply and associated supporting hardware.

Small cells additionally require a different complexity, particularly around SON [Self-Organising Networks]. This aspect is something that the femtocell industry has had to get right in order to scale up to the millions of units in commercial use today. The ability to identify neighbour cells and determine the right frequency, codes and RF power level to operate at are essential to reduce the OPEX costs. Simple plug-and-play capabilities are also important to reduce installation time, skillset required and cost.

What’s the profile of your LTE and 3G small cell customers today?

We now have a combined total of 55 customers today who have developed products ranging from residential, enterprise through to pico/metro. What’s changed is that in the second half of 2011 we have seen demand for dual-mode 3G/LTE small cells. This is especially coming from the Far East – specifically Japan and Korea.

Perhaps surprisingly, these combined 3G/LTE products are being initially targeted for residential use. In these countries [Japan and Korea], high speed wireline broadband is very common – it’s not unusual to have 100Mbit/s broadband at home – which can easily support the high speeds this technology is capable of.

Whether these products are used outdoors and in public areas first in other regions, only time will tell.