Brian Antalek has given a nice explanation explained it nicely in his paper "Using pulsed gradient spin echo NMR for chemical mixture analysis: How to obtain optimum results" results"
( http://www3.interscience.wiley.com/journal/94519618/abstract )

There is additional information in "Improving the accuracy of pulsed field gradient NMR diffusion experiments: Correction for gradient non-uniformity", by Conell et al. .
( http://dx.doi.org/10.1016/j.jmr.2009.01.025 )