Distorted signal recovery (Technion)
- Summary
- Non linear response, that distorts samples of continues signals, is a common phenomena in acquisition devices such as sensors, detectors and analog to digital converters (ADC). Two main approaches are used when reconstructing continues time signals from such samples: the first assumes a linear response of the acquisition system. This assumption is often not valid in practice. The second approach assumes instantaneous non linear response, and ignores any memory effects caused by the sampling system. Our approach takes into account the nonlinear response and memory effects of the sampling system, allowing our method to be applicable to a wide range of cases.Our algorithm iteratively processes the available samples, in order to reconstruct continues signal. We impose conditions to assure convergence of the algorithm and show that the solution found is unique and allows perfect reconstruction. The method works for numerous signals and systems, and in practice, avoids any assumptions and constrains. The algorithm makes linear approximations of the sampling scheme, in order to reconstruct the input, then computes the samples of such input by the non linear acquisition device and iterates the result again till convergence is reached. Our approach is based on a proven mathematical theory and is tested successfully.
- Technology Benefits
- Perfect reconstruction of distorted signals
Suitable for a wide range of signals
Applicable to non-ideal scenarios
Practically of no limiting assumptions
- Technology Application
- Improved communication systems
Cameras with increased dynamic range
Optical sampling systems
- ID No.
- COM-0970
- Country/Region
- Israel
For more information, please click Here
