Efficient Design and Implementation of FIR filters on FPGA for Wireless Networks

Abstract

With the advent of Internet of Things (IOT) and Industrial IOT (IIOT) there is always newlinea need for an efficient digital processing technique for an analog signal generated. A newlinelot of sensors are required to extract the analog signals. Currently, the digital filters newlineare inadequate with excess computational cost, latency and hardware utilization. The newlineresearch in this work is an attempt to provide a solution to the above issues by achieving newlinean effective design of filters through the developed hybrid optimization algorithm newlineby combining Brain Strom Optimization (BSO) and Artificial Bee Colony Algorithm newline(ABC). The developed algorithm is implemented on Field Programmable Gate Arrays newline(FPGA) and evaluated for the performances like latency, number of Look up Tables newline(LUT) along with computation of Mean Absolute Error (MAE). Also, the developed newlinealgorithm is compared with standard algorithms such as Particle Swarm Optimization newline(PSO), Artificial Bee Colony algorithm (ABC) and also with different algorithms in newlineMATLAB environment with the implementation in FPGA. These performance measures newlineplay a key part in the design and implementation of optimal digital filters. With newlinethe advancement of signal processing applications, the design of digital filters has received newlineconsiderable attention due to the widespread applications in numerous fields like newlinecommunication network, power system network, and multidimensional signal processing. newlineHence, this research work addresses the issues in base-band processing section of newlinewireless receiver and propose a three techniques to eliminate the noise contained in the newlinesignal using the optimization-driven techniques. The focus in this research is on both newlineDecimation filter and Farrow Structure-based FIR filter. One of the main objectives newlineis on how the BSO is modified by ABC to update rule. The implementation results newlineproved that there is reduction in rate of LUTs approximately by 7 % and power gain newlineby 19 % for a 16-point filter with the execution time of 9.129 ns which is smaller than newlineother techniques. The Com

Description

Keywords

Citation

item.page.endorsement

item.page.review

item.page.supplemented

item.page.referenced