Wavelet transforms as they apply to optimal receiver design are studied. We start with an overview of the Karhunen-Loéve transform and explore the relationship between wavelet bases and the Karhunen-Loéve transform. We show that the dyadic wavelet basic can constitute the eigenfunction basis of a random process. With the help of this foundation, the design of an optimal receiver by the use of a the wavelet expansion is described. The relationship of this receiver to the wavelet-transform-based adaptive filter is also established.