Optimal time and frequency domain waveform design for target detection

2010

Oceans'10 Ieee Sydney

Self Cite

The design of transmit waveforms in sonar and radar that optimize the probability of detection of a known target is an area of active interest. Previous research has demonstrated the benefits of optimal transmit waveforms for enhanced target detection, versus the transmission of more conventional waveforms such as broadband LFM pulses. For maximum benefit, the scattering function or frequency response of the target to be detected must be known, along with the spectral or statistical properties of the environment. In this paper, we examine the impact of a mismatch between the expected frequency response of the target, for which the optimal transmit waveform was designed, and the actual frequency response of the target. In particular, we design the optimal sonar transmit waveform based on the free-field response of the target, and then examine the effects on detection performance of burial of the target in sediment. Simulations of the sonar backscatter from a steel sphere in the water column and at various stages of burial demonstrate that the impact of mismatch becomes increasingly detrimental as the target becomes increasingly buried. For fully buried targets, the impact is such that transmitting a simple broadband LFM pulse can yield improved detection relative to "optimal" waveform design wherein there is a mismatch between the expected (free-field) and actual (buried) frequency response of the target.

show abstract

Section: Discussionmentioning

confidence: 98%

Section: Background: Optimal Signal Designmentioning

confidence: 99%

See 1 more Smart Citation

Sonar waveform design for optimum target detection: The impact of object burial state

2010

Oceans'10 Ieee Sydney

Self Cite

show abstract

“…In our work we formulate the classical concentration problem as a constrained nonlinear optimization problem, to which we add constraints that incorporate the optimal spectral magnitude from [4,5,13]. Solving this modified optimization problem produces a time-domain signal that is real, maximally concentrated in the discrete-time interval (0, N − 1), and has a magnitude spectrum that is arbitrarily close, in the least squares sense, to that which maximizes detection performance.…”

Section: Introductionmentioning

confidence: 99%

“…In particular, we consider transmit waveform designs based on two problem formulations. In the first formulation [13], we directly design spectral phase functions based on the desire to minimize or maximize the duration of the waveform, subject to the optimal spectral magnitude criterion developed by Kay in [4] for point targets and extended to elastic targets in [5,13]. The resultant waveforms give the designer the freedom to choose signals with short duration but also high peak energy or signals with lower peak energy and longer duration, while maintaining optimal detection performance.…”

Section: Introductionmentioning

confidence: 99%

Waveform design with time and frequency constraints for optimal detection of elastic objects

2011

SPIE Proceedings

Time and frequency constrained sonar signal design for optimal detection of elastic objects

The Journal of the Acoustical Society of America

2013

Self Cite