“Which patent is you most proud of?”, is also a question I have got several times. Among all 400+ filed it is a hard question answer, since it depends on what you mean. Some inventions are about really clever details requiring a lot of research to figure out, while other patents, still new and inventive, does not require “rocket scientist” education to figure out the solution, but are important from an business perspective.
But if I have to choose, one of my favorites from a cleverness point of view is (I’m sorry for the technical details, below)
US6725024 Offset local oscillator frequency
I was visiting Ericsson Research in Rayleigh, NC, US a week end of summer 2000, and discussed with some of my research colleagues there, about improving the performance of the channel estimator for the highest data rates in EDGE. It turned out that, with a low cost direct conversion receiver, the DC component introduced by the receiver and channel estimator severely limited the performance of EDGE in the mobile device implementation. Is it a way to achieve better estimation performance by doing something in the radio receiver? After some hours of discussion, we found the answer. By offsetting the local oscillator in the radio with some few kHz, different offset for different GSM/EDGE training sequences, channel and DC offset performance coming close to the Cramer-Rao bound (a measure on ho how good a channel estimate could be) can be achieved. We later on presented the results at a Conference as well.
- Hui, B. Lindoff and K. Zangi, Enhanced DC estimation via sequence-specific frequency offset, In the proceedings of 56th IEEE Vehicular Technology Conference (VTC 2002 fall), Vancouver, Canada, September 2002, p161-165, vol.1.