Title :
Application of parallel computing technology OpenMP to search for the generator polynomials
Author :
Mytsko, Evenly ; Malchukov, Andrey
Author_Institution :
Dept. of Comput. Eng., FSAEI HE NR Tomsk Polytech. Univ., Tomsk, Russia
Abstract :
The paper deals with the application of parallel computing technology OpenMP to search for the generator polynomials. Forming the task of finding polynomials to create polynomial block codes is quite laborious. For example for generator polynomial, on which basis the fail-safe code (m = 32, t = 3), you need two weeks of continuous work program on the processor Intel XEON 5150 with a core clock of 2.66 GHz and RAM 1 GB. The article describes the search algorithm forming polynomials, which are constructed on the basis of codes more efficient than Bose-Chaudhuri-Hocquenghem codes. The article describes the ability to form the polynomials and analyzed in terms of the possibility of using parallel computing technology OpenMP. Computer experiment was delivered to compare the performance of software implementations with and without the OpenMP. Results supplied by the computer experiment to study the performance of sequential and parallelized implementations of the search algorithm polynomials showed that the usage of OpenMP technology allows under certain input parameters (m = 9, t = 2) to increase the speed of 20-24 times. On average, using of OpenMP algorithm enhances performance up to 3 times.
Keywords :
block codes; error correction codes; message passing; parallel processing; polynomials; search problems; Intel XEON 5150 processor; OpenMP technology; error correcting code; fail-safe code; generator polynomials; parallel computing technology; polynomial block codes; search algorithm polynomials; sequential implementation; Acceleration; Random access memory; Uniform resource locators; Hamming distance; error-correcting code; generator polynomial; parallel computing; search algorithm; threads;
Conference_Titel :
Mechanical Engineering, Automation and Control Systems (MEACS), 2014 International Conference on
Conference_Location :
Tomsk
Print_ISBN :
978-1-4799-6220-4
DOI :
10.1109/MEACS.2014.6986902