NXP Semiconductor reckons that GPS in cellphones is going to be a very big deal That’s why it bought fabless semiconductor GPS specialist company, GloNav, last month.
“GPS is really core to our strategy”, says Mark Cetto, Executive Vice President for the Mobile and Personal business unit at NXP.
The reasoning went like this: “We were not on a path internally to intersect with our timetable”, says Cetto, “the internal approach was based on software. But software-based GPS is not suitable for cellphone applications because it needs a lot of MIPS, and because the time to the first fix is much quicker in hardware then in software.”
Furthermore: “The RF we were working on is BiCMOS but we needed RF CMOS to integrate the GPS”, explains Cetto, “doing GPS in RF CMOS is more difficult than in BiCMOS. To integrate GPS RF with the GPS-baseband to do a single chip GPS is very complicated. A GPS receiver is a very sensitive receiver. If you put it on the same die as high speed digital logic, the noise of the high-speed digital logic damages the receiver. It’s much more difficult than a single chip Bluetooth because of the level of sensitivity. Bluetooth is a millions of times more powerful signal.”
Those were the problems. What the GloNav acquisition brings to the party is, first, it already has a single chip RF CMOS GPS in 90nm technology; and secondly, it has 20 years experience of selling GPS.
“The GloNav GPS solution is mature”, says Cetto, “it has been deployed in the field for 20 years. We know It can deal with reflection and urban canyons.”
NXP’s task is now threefold:
First: to incorporate the GPS into Nexperia, NXP’s wireless platform. That will be done this year.
Second: the silicon integration of GPS in the cellular baseband. NXP expects to be sampling this early next year, probably on 45nm silicon.
Third: work on connectivity combos. This will be done in parallel with the integration of the GPS into the baseband.
Subscribe to this blog's feed