Simulation of HSTL IO standard based energy efficient Punjabi Unicode reader on FPGA

摘要

Punjabi is ranked 1 in languages of Pakistan [5], 11th in Indian languages [6] and 3 in Indian Subcontinent. In order to write Punjabi: Gurmukhi script is used in India and Shahmukhi in Pakistan. A lot of research is going on in Pakistan and India in the domain of natural language processing but, no research group is working especially for Punjabi to design Unicode reader. The Unicode range of characters for Gurmukhi script is 0x0A01-0x0A75. In this work, we are trying to fill this research gap in Punjabi natural language processing. This paper covers the hardware design and implementation of Punjabi Unicode Reader (PUR) for Gurmukhi scripts on Virtex-6 FPGA. This PUR design was tested with device operating frequency of 25 MHz, 125 MHz, 625 MHz, 1 GHz and 25 GHz. HSTL (High Speed Transceiver Logic) and I/O standard is used to make this design more energy efficient. Using HSTL_I_12, 38-50% I/O power was noticed and while using HSTL_II,the saving increases to 49-61%, with the device operating in range of 25MHz to 25GHz.This Unicode reader can count the number of vowel, consonants, and digit when used with counter.