Tuesday, March 15, 2016

Nyquist in reverse

In signal processing, the Nyquist rate, named after Harry Nyquist, is twice the bandwidth of a bandlimited function or a bandlimited channel. This term means two different things under two different circumstances:
  1. as a lower bound for the sample rate for alias-free signal sampling[1] (not to be confused with the Nyquist frequency, which is half the sampling rate of a discrete-timesystem) and
  1. as an upper bound for the symbol rate across a bandwidth-limited baseband channel such as a telegraph line[2] or passband channel such as a limited radio frequency band or a frequency division multiplex channel.


OK, this is from Wiki:

In point 1, you are allowed to confuse alias free sample rate and Nyquist frequency of discrete things, the two are closely related.

The main  point is riun  Nyquist in reverse. If, say, the oil consumer sees a price peak, exceeding the previous price peak, then a spectral constraint must be met.  The consumer has two samples, and therefore the sequence of pricing events between the perks is a representative sample.

For example: Oil back end flow is calibrated to the representative sample..

So the oil frackers,they get it,they develop optimum shut down and restart technology they can do the representative sample.

It's the deja vu effect. Deju vu effects occur all the time and cause local reordering of sequences.  The sequence is generated from a probability graph, like the ones you get with Huffman compression. .

A bit of tautology
I think the derivation through equipartition means this is a bind because the topic is price, which requires a quotient field.  Leading to the more fundamental, why are quotient fields entropy maximizing? Is Ito's calculus a necessity of natural things?

No comments: