Title :
New algorithmic technique for linearizing diode detectors
Author :
Nijhawan, Garima ; Hashmi, M.S.
Author_Institution :
Landis+Gyr Noida, Noida, India
Abstract :
This paper presents a linearization approach to extend the dynamic range of diode based power detector. The linearized power is estimated using support vector regression with the regularization parameter estimated directly from the training data. The proposed technique gives 25% better results as compared to previously published results. Comparison of diode power detector characteristics for different excitations is also done.
Keywords :
diodes; linearisation techniques; power measurement; regression analysis; support vector machines; algorithmic technique; diode based power detector; diode detectors; linearization approach; linearized power estimation; regularization parameter; support vector regression; Detectors; Dynamic range; Linearization techniques; Schottky diodes; Support vector machines; Table lookup; Training data;
Conference_Titel :
Multimedia, Signal Processing and Communication Technologies (IMPACT), 2013 International Conference on
Conference_Location :
Aligarh
Print_ISBN :
978-1-4799-1202-5
DOI :
10.1109/MSPCT.2013.6782112
