A Simulator for Depicting and Comparing Adaptive Algorithms in Signal Processing

[attachment=18008]

I. INTRODUCTION
In this fast moving world, it is very difficult and time consuming for the user to observe the things separately for making a decision regarding his requirement. For making the user requirement in a simple and most friendly way in the area of adaptive filters, (even here the user has a problem in deciding which algorithms he/she should use for an application) we came up with a simulator in which the user can view and compare the output of a different algorithms for his requirement and can choose among the best. A simulator has been designed to make his requirement in a very easy way.
In this project a study of all the algorithms like Least Mean squire (LMS), Normalized Least Mean Squire (NLMS), Leaky Least Mean Squire (LLMS), Recursive Least Squire (RLS), Sign Least Mean Squire, Sign Normalized Least mean Squire has been performed. The performances and results of each application like Adaptive Line Enhancer, Adaptive Noise Canceller and System Identification by applying each algorithm can be performed using the template shown in the Figure. 1. Matlab Graphical User Interface (GUI) [4] is used for designing the template. GUI is an user friendly interface where the user can observe the desired signal, estimated signal, output signal, convergence of coefficients and learning curve in a deigned window by entering the desired values and selecting the algorithm and application. The basic application structures are introduced and some of the classical "real-world" applications linked to these applications.



III. APPLICATIONS
A. Adaptive Noise Canceller
One of the most common practical applications of adaptive filters is noise cancellation. The Application of Adaptive Noise Canceller (ANC) is to remove the noise from signal. The concept of Adaptive Noise Cancelling, an alternative method of estimating signals corrupted by additive noise or interference. The method uses a ‘primary; input containing the corrupted signal and a reference input containing noise correlated in some unknown way with the primary noise. The reference input is adaptively filtered and subtracted from the primary input to obtain the signal estimate [1]. The structure of an ANC is illustrated in Figure. 2 and Figure. 3 illustrates ANC using LMS algorithm.




V. CONCLUSION
In real time signal processing applications the part played by adaptive filters are requisite. Opting most beneficial algorithm for the application plays a crucial role in system performance. For this reason a study of Adaptive algorithms has been performed and the performance of each application can be observed by using each and every algorithms mentioned in the simulator. Regarding the pros and cons of each algorithm is observed by using simulator. Using the simulator we prepared it becomes handy for the user to compare the outputs of different algorithms and can choose the best among them according to his use.