But for t ^ 0, (SNR)c = BT/n. Another SNR, (SNR)t, was found to be
(SNR)t - Nt ^/VARjx+} = NbTb2 Pxy (t)/[oy(1 - pxy(t))] 8.2-38
where T is the total observation time over which spikes are processed, and pxy(T) is defined as the standardized cross-correlation function, given by
Pxy(t) = Rxy(t)/(oxOy) = ^(2B/rai) [sin(BT)/(Bx)]cos(ffl0t) 8.2-39
i for the IBPF. Wu goes on to define a "noise figure," F, for TC applied to the IBPF.
At t = 0, the noise figure reduces to
2G2lo(. - 2B) r 2 . 2 F = 2 I, '2 w exp[b2V4G\B] 8.2-4l n by (12Q2 +1)/6
As a performance index, F is desired to approach unity. Thus, the TC process gives the best results when the IBPF Q is large. Interestingly, too large a noise bandwidth, rai, works against one in making F large. However, rai must be large enough so that the estimate of Rxy(x) is a valid representation of h(x), the linear system weighting function. Common sense dictates that, in any case, rai must be > (rao + B).
Was this article helpful?
This guide will help millions of people understand this condition so that they can take control of their lives and make informed decisions. The ebook covers information on a vast number of different types of neuropathy. In addition, it will be a useful resource for their families, caregivers, and health care providers.