Return Of The Femtocell

By Cheryl Ajluni

Nothing has been left unscathed in the current global economic downturn, and that includes femtocell deployments.

It was just last year that femtocells were being proclaimed a 2009 “killer app,” along with LTE and WiMAX. But what was once viewed as the next great thing has instead faced a tough road with more than a few large-scale deployments by major mobile operators being put on hold. Despite the slowdown, femtocells remain a viable part of the road to 4G. According to Aditya Kaul, senior analyst at ABI Research, while the pace of adoption has been slow, “deployments in 2010 will pick up.” In fact, ABI Research forecasts that the total available femtocell market in 2010 will reach 2.3 million units and rise to 40 million units by 2014.

Adding fuel to the slow-burning fire of the femtocell market is movement on the part of carriers, which may signal that things are heating up once more. Comcast, for example, announced that it is testing WiMAX femtocells. And across the globe, China Mobile has partnered with Nokia Siemens Networks to test a TD-LTE femtocell. This year, six major network operators have launched femtocell services that cover the USA, Europe and Asia.

Such activities bode well for the emerging femtocell market, but do little to address designer’s challenges at a system level. From a technical perspective, femtocells are no more complex than a full macro base station, yet they demand the integration level of a WLAN in order to meet various cost, power and footprint requirements. And, since femtocell products are likely to appeal to many different users around the world, each with potentially different needs and requirements, different models will need to be developed. Each model (e.g., a W-CDMA, WiMAX or LTE femtocell) will have its own unique design requirements and challenges determined by the standard it supports. Such challenges make designing femtocell products a difficult and risky proposition, and demand solutions that can ease the burden on the system designer.

As Caroline Gabriel of Rethink Research, explains “The femtocell market is starting to mature. While the sophisticated early entrants wrote their own solutions, this is bringing an inevitable demand from developers for more complete system solutions that reduce risk, enable differentiation and speed time-to-market. This is a value chain dynamic that other semiconductor markets such as Wi-Fi and DSL also experienced as they hit the mass-market.”

picoChip is working on just such a solution—an end-to-end femtocell reference solution for both HSPA and LTE that’s aimed at tackling the designers’ system-level challenges, including interference management, security, timing, and provisioning. The solution integrates optimized versions of Continuous Computing’s protocol stacks with picoChip’s SoC products (Figure 1). It also includes sophisticated femtocell management software for the complex control functionality required by femtocells.

Figure 1. The four-user PC302 residential femtocell is the first in picoChip’s PC3xx family of highly integrated baseband SoCs. It implements a complete 3GPP Release 7 access point, is compliant to TR25.820 and the newly standardized Iuh interface, and is built using an advanced 65-nm manufacturing process.

Of course, even the availability of a complete femtocell solution with things like automated interference management and network self organization, does not exempt the designer from having to consider some fairly difficult issues during femtocell design. Some of these issues include:

Power Consumption: By definition, femtocells are low-cost, low-power wireless access points that operate in licensed spectrum to connect standard mobile devices to a mobile operator’s network using residential DSL or cable broadband connections. Power is an especially critical concern since in a residential situation the end user has to pay the bill for the electrical energy consumed by the femtocell base station. Its power consumption therefore, has to be low enough to not significantly impact the user’s bill. Various low-power design techniques such as comprehensive power-down modes and clock gating, as well as effective power management can be employed to achieve this goal.

Interference: Because femtocells utilize spectrum currently employed by macro networks, interference can result between cells (e.g., a macrocell with a femtocell or a femtocell with another femtocell, when two units are in close proximity). Interference affects operators as well as consumers who will likely move their femtocells around or place them next to other devices.

There are a couple of ways to mitigate the risk of interference. The first, cognitive radio technology, essentially allows the femtocell’s radio to constantly monitor its RF environment and set itself up accordingly. Another option calls on the femtocell to intelligently set its output power when it’s in the presence of nearby femtocells. In this case, each femtocell would need to transmit at lower power to avoid same-frequency interference. To avoid interference with signals from neighboring macrocell base stations operating on an adjacent channel, the femtocell can be designed to measure the power in the adjacent channel downlink and set its own power accordingly.

Standards: These are key to taking a technology from a niche application to wide-scale adoption. The question with regard to femtocell technology is which standard—a proprietary approach, an existing standard from the telecom industry (e.g., the session initiation protocol (SIP)/IP multimedia subsystem standard for integration used in LTE networks), or some other standard altogether?

The Femto Forum, 3GPP and the Broadband Forum think they have the answer. It’s the world’s first femtocell standard and they created it together. The standard, part of 3GPP’s Release 8 and interdependent with Broadband Forum extensions to its Technical Report-069 (TR-069), was officially released in April and paves the way for not only the development and production of large volumes of standardized femtocells, but also for ensuring interoperability between different vendors’ access points and femto gateways. It covers four main areas: network architecture (the Iuh), radio & interference aspects, femtocell management/provisioning via the popular TR-069 protocol, and security (e.g., IKEv2 and IPsec).

Handovers: These can be tricky, especially in the case of femtocells where precise timing and synchronization is need to properly manage handovers with the macro network. The process is complicated by the fact that there may be millions of femtocells deployed and end-users may move them around their homes, entering into areas where the signal strength from the macrocell is greater than that of the femtocell. Unfortunately, existing macrocell RF planning techniques offer no real solution. Instead, femtocell handovers require sophisticated algorithms capable of ensuring that the network quality is not impacted by inefficient handovers and wasted capacity.

These issues are just some of the challenges system designers face when developing femtocell products today. Addressing them, whether by using the techniques suggested or through a complete femtocell solution like the one picoCell is working on, will be critical to ensuring not only successful femtocell products, but a successful femtocell industry that can take its rightful place on the road to 4G.