Extracting Signal Information from Contaminated Data with a Two-Dimensional Array.
NAVAL UNDERWATER SYSTEMS CENTER NEW LONDON CT NEW LONDON LAB
Pagination or Media Count:
The use of sensors configured into various geometrical shapes called arrays are important in extracting signal information in many applications. This paper explores the effectiveness of a two-dimensional or planar array in extracting signal information from digitized data when additive Gaussian noise and additive impulsive interference are present. Based on the likelihood ratio approach the data output of each sensor is first passed through a zero-memory device. The zero-memory device is neither a nonlinearity or linearity depending on whether the interference is present or not present, respectively. Further processing utilizes a receiver which converts the data at the output of each zero-memory device into frequency components using a discrete-Fourier transform. The frequency components are then weighted and summed over all spatial sensors. The weights are chosen by maximizing a defined performance measure. It is shown that this optimization procedure does not take into account the debilitating effects of impulsive interference. In order to treat theoretically the inclusion of a nonlinearity, to combat impulsive interference, the concept of an ideal nonlinearity is introduced.
- Numerical Mathematics