Optimal invariant detection of a monochromatic plane wave with unknown amplitude, frequency, phase, direction of arrival and noise variance

Enhanced beamspace MUSIC for cost‐effective FMCW automotive radar

“…A reasonable beamformer design is important for beamspace MUSIC, e.g. improving the parameter estimation [19] and the resolution [20, 21]. Many beamformer designs have been proposed for a higher resolution or smaller mean square error and their performance is evaluated via simulations given an extremely large number of data samples [22–25].…”

Robust DFT‐based generalised likelihood ratio test for underwater tone detection

“…Generally, there is no prior information of the tonal signals and under some reasonable assumptions, the ocean ambient noise usually follows a Gaussian distribution [3,4]. Therefore, taking the statistical characteristics of the random noise into consideration, the problem of detecting a tone with unknown parameters contaminated by Gaussian noise with unknown variance can be formulated as a problem of composite hypothesis testing [5][6][7][8][9][10].…”

“…For the problem to be discussed, the uniformly most-powerfulinvariant (UMPI) detector does not exist. However, given the signal-to-noise ratio (SNR), the UMPI performance bound can be computed and used as an upper bound on the detection performance for any invariant test [9][10][11]. A generalised likelihood ratio test (GLRT) for tone detection is derived in [10], whose performance approaches the UMPI performance bound as the SNR approaches infinity [8,10].…”