IOSR Journal of VLSI and Signal Processing (IOSR-JVSP) Volume 6, Issue 3, Ver. I (May. -Jun. 2016), PP 31-37 e-ISSN: 2319 – 4200, p-ISSN No. : 2319 – 4197 www.iosrjournals.org DOI: 10.9790/4200-0603013137 www.iosrjournals.org 31 | Page Reconfigurable Distributed Arithmetic Based Adaptive Noise Canceller Using Modified NLMS Algorithm Dr. Rajesh Mehra 1 , Lalita Sharma 2 1, 2 Department of Electronics & Communication Engineering, NITTTR Chandigarh, Punjab University, India. Abstract:This paper presents an efficient design and implementation of low area, high speed Adaptive filter based on Distributed Arithmetic (DA) Scheme. An enhanced NLMS algorithm has been proposed for the adaptive noise cancellation filter design. The computation speed of the proposed NLMS system is relatively high due to preallocation of memory for variables in enhanced Normalized LMS algorithm.The proposed design is successfully implemented using Matlab Code and Xilinx ISE Design Suit on Spartan 3 based XC 35400 and Spartan 3E based Xc3500e FPGA device. The synthesis report shows a considerable decrease in device utilization percentage and increase in overall speed than the existing design. For 20 tap proposed filter there is 43% reduction in number of slices, 59% reduction in number of flip flops, 24% reduction in number of LUTs used, whereas 54% improvement has been achieved in maximum frequency and 35.14% improvement in minimum period. Whereas for 10 coefficient filter there is 21% increase in maximum frequency and 16.46% decrease in minimum period. Keywords: Adaptive Filters, Distributed Arithmetic, FPGA, NLMS Algorithm, Noise Cancelation. I. Introduction In this era of extensive telecommunication systems, the efficient signal processing is one of the biggest challenges as the signals suffer interference and noise caused by various transmission mediums. To improve the quality of communication, an effective noise cancellation method is required [1]. Noise Cancellation refers to the process of optimal filtering that includes estimation of the noise by filtering the reference signal and deducting this estimated noise from the primary input which contains both signal and noise. In adaptive filtering process when an input signal containing noise is applied to the filter, a negative feedback is applied which depends on the noise in the input signal by adjusting weights values which cancels out the noise from input signal [2]. Over the past two decades, digital signal processors have been changed revolutionary to improve speed and efficiency of communication systems. Many advancements have been made in DSPsover the past three decades in speed improvement, area and power consumption. The researchers have put a great effort in crafting efficient Digital Signal Processing (DSP) functions architecture such as FIR filters, which are most commonly used in various telecommunication applications. Adaptive filter is one of the effective solution to filter out noise less signals in a communication system.Adaptive filter changes filter coefficients with time to adapt to the dynamic input signal environment [3]. Fig. 1 represents a standard Adaptive Noise Cancellation process in which there are two inputs: one is primary signal and other is reference signal. The primary signal x n is corrupted by the noise n n added by means of communication mediums or external environment. The reference signal nr n is second input which is similar to or correlated with the noise signal n n . The reference noise passes through an Adaptive Filter to produce an output nf n which nearly resembles noise n n present in the primary input [4]. This estimated noise ( nf n ) is subtracted from the primary input signal (x n + n n ) to produce the estimated error e n and output y n which is similar to the signal x n . Fig. 1: Standard Adaptive Noise Canceller Organization