Lower-bound on SNR of the nth-order hyperbolic time–frequency kernel

Abstract

An estimate of the lower-bound on signal-to-noise ratio (SNR) of the nth-order hyperbolic time-frequency kernel is given. The effects of kernel parameters such as ߬ t, n and a on the SNR are discussed. In particular, the direct relationship between the SNR and auto-term slope a is studied in detail. Conditions under which the lower-bound on SNR is obtained are derived. Preliminary observations on a transfer function model with the auto-term slope a and ߠas inputs, and lower-bound on SNR as output are given. Possible further work is outlined.