An Optimization in Conventional Shift &Add Multiplier for Area-Efficient Implementation on FPGA

Abstract

FPGA is familiar with prototyping and implementing simple to complex DSP systems. The FPGA-based design may be highly affected by factors that include selection of an FPGA board, Electronic Design Automation Tool and the Programming Techniques to optimize the algorithm. The algorithm optimization results in a more compact design regarding the area and achieved frequency.In DSP algorithms optimization, the major bottleneck is the multiplier complexity evident in, for example - FIR, IIR, FFT, and others. Research shows much work on multiplier optimization. Despite all possible optimization techniques, the multiplier consumes tremendous resources when translated on hardware, with more power consumption and observed delay. The proposed work is novel in that it brings resources optimization in a familiar shift and add multiplier algorithm by implementing the design in FPGA and comparing the results with the existing shift, and add a multiplier. In the implementation of the design, Xilinx Vertex −7 FPGA is used along with ISE 14.2 simulators. The parameters to compare are the Lookup tables (Logic element of FPGA), adder/subtractors and the multiplexers, along with performance characters, like the operating frequency, delay and total levels of logic(path travelled by the signal in register transfer level). The output shows that the anticipated design is an excellent alternative to the conventional shift and add algorithm.