Using Generalized Entropy Approach for Blind Estimation of the Length of Spreading Sequences in Code Division Multiple Access Systems

In this paper, a novel method for signal detection and blind estimation of the length of spreading sequences, in direct sequence code division multiple access (DS/CDMA) systems with different short and long code modes, is presented. The spreading sequences are the basis of the spread spectrum systems which is mainly modeled as a quasi-random process, and their identification and estimation is the most important part of the structure of a spread spectrum receiver and the present article addresses this important issue. The basis of the proposed method of code length estimation is the information theory learning (ITL), and two of its major categories namely entropy criterion and mutual information criterion, which are used in modeling stochastic processes. The proposed method is a kind of generalized entropy and because it includes the integral of the power of probability density function, it is easier to estimate than the Shannon entropy and in large scale implementations it shows better performance. Besides, the proposed method does not need the basic information and that is why it is used in the design of blind receivers. Simulations and numerical results show the validity of our claim in the ability of the presented method in the estimation of the length of spreading sequences in both short and long coding modes.