I spoke with Nicholas Karter, Senior Director, Business Development at Qualcomm, to uncover the details of their femtocell chipset development program. Here’s a quick update on what they’re doing.
What is Qualcomm planning to deliver?
The Femto Station Modem (FSM) will be commercially available before the end of 2010. There will be two versions in their first release:
- 9800 for CDMA
- 9200 for UMTS/HSPA
The CDMA chipset will support both 2G and 3G flavours, known as 1xRTT, EV-DO Rev A and Rev B. Most CDMA network operators, such as Verizon, are using EV-DO Rev A today, although some such as MetroPCS has kept with 1xRTT. The first EV-DO Rev B network went commercial at SmartPhone in Indonesia in January 2010.
The UMTS chipset will support 3GPP Release 7, including HSPA (High Speed Data Access) rates up to 28 Mbit/s using 2x2 MIMO.
In some cases, this means that femtocells based on these chipsets may exceed the performance of the outdoor macrocell network. Some service providers are evaluating if this is a good idea or not. [Ed Note: I would think there would be more benefit and this approach is also true of other femtocell chipsets]
Achieving these peak speeds also relies on having the latest handset device, compatible with these high specifications.
When will it be available?
Both engineering programs are on schedule, with the chips in various stages of test. Commercial availability is slated before the end of 2010.
What unique features and capabilities are incorporated?
The chipset is designed as a one stop shop/fully integrated solution, including A/D conversion, RF front end, GPS, timing inc. IEEE 1588 v2, network listen and power management.
Interference between femtocells and the macrocellular network has been a concern for network operators, although several have gone on record to say that the issues are manageable. Qualcomm contributed to a study report on the subject published by the Femto Forum, which made a number of recommendations to mitigate issues in difficult cases. These and other interference management features have been designed into the product.
GPS is also integrated (mandatory for US carriers, and of interest elsewhere for location services). Only a single low cost oscillator crystal is required (other designs often require separate crystals for GPS and the main baseband processor).
Power management is also fully integrated onboard.
Is there any software included?
Qualcomm’s strategy is not to “go up the stack” and won’t be offering higher layer protocol or application software.
It is expected that femtocell vendors will reuse their existing software from other hardware designs, or acquire software from other vendors to use with it.
Who will buy it?
There are two public announcements of customers already – ZTE and Airwalk – for the CDMA version.
The price point for the overall hardware platform (the BOM or Bill of Materials) will be competitive in the market.
The RF chipset is “LTE Ready” and able to support the wider bandwidths and MIMO required by this new 4th Generation technology. This suggests that future releases may include LTE as an alternative or even as a combined 3G/4G multi-mode solution.
My take on this
Qualcomm should clean up on the 3G CDMA femtocell chipset market. There isn’t a dedicated System-on-a-Chip design for CDMA available today, and this combined with Qualcomm’s renowned expertise in the technology (they practically invented it) should give them substantial market dominance. picoChip and Percello don’t address this technology today, believing that UMTS and LTE markets hold greater prospects.
The market position for UMTS/HSPA will be more difficult to establish. The dominant market presence from two strong encumbents (picoChip and Percello) presents a greater challenge to this new entrant.
I look forward to hearing more press announcements from Qualcomm about their platform.