# Determining bits to ensure a particular quantization level in signal sampling (DSP)

1. Sep 8, 2012

### peterpiper

1. The problem statement, all variables and given/known data

Consider the following discrete-time signal where the samples are represented using N bits.

x(k) = exp(-ckT)μ(k)

μ(k) represents the unit step function and T is the Δ between each sample.

-How many bits are needed to ensure that the quantization level is less than .001?

2. Relevant equations

q = $\frac{x_{max}-x_{min}}{2^{N}-1}$

3. The attempt at a solution

I have yet to find a way to even attempt this solution. Is there a common range of x values that are used in a sampling process in order to find the appropriate quantization levels or is a range even necessary given the generalized signal representation?

I'm using Fundamentals of Digital Signal Processing using Matlab and it has been no help as far as I've seen. If anybody has used this textbook in the past and point me towards some useful resources it'd be greatly appreciated.