Adaptive signal processing design and implementation
MetadataShow full item record
In an effort to produce an efficient means of applying adaptive filters to practical application, the Adaptive Signal Processing Design and Implementation (ASPDI) interactive package has been developed. ASPDI includes signal generation, file management, adaptive filters simulation, signal analysis, graphics, and TMS320C25 digital signal processor assembly code generation facilities. ASPDI allows repeated evaluation of the performance of a filter in order to select optimum structure, algorithm, and parameter values for a particular application. The signal generation, graphic, file management, and signal analysis compose an environment for data acquisition, analysis and display. The adaptive filters simulation supports four structures: Transversal, Lattice, Infinite Impulse Response (HR), and symmetry transversal, and eight algorithms: Least Mean Square (LMS), Leaky LMS, Normalized LMS, Sign-Data LMS, Sign-Error LMS, Sign-Sign LMS, Kalman, and Recursive Least Square (RLS). Many fundamental C function routines that provide rudimentary capabilities for developing new structures and algorithms for adaptive filters are included. The code generator produces the most efficient assembly language code output for the TMS320C25 digital signal processor by modifying an appropriate source file according to the parameters input by the user. The subroutines which can be called by TMS320C25 assembly or C language main program for adaptive filters of various structures and algorithms are also developed in this thesis.