Average Likelihood Method for Classification of CDMA

A New algorithm for classifying Direct Sequence Code Division Multiple Access (DS-CDMA) signals in additive white Gaussian noise (AGWN) is proposed based on the average likelihood (AL) function. The AL is express in terms of the code length and the number of active users. These parameters constitute the hypothesis under test. While the AL function is expressed in an analytical form, the implementation requires knowledge on the spreading codes that achieve a low Total Squared Correlation (TSC) value. This complexity can be simplified by developing a hard decision scheme which only requires knowledge on matrices that achieve the lowest TSC. Our study cases include binary and multiclass classification using full loaded DS-CDMA with knowledge on the chip period. Code lengths of powers of 2 are used as a proof of concept.