• Àüü
  • ÀüÀÚ/Àü±â
  • Åë½Å
  • ÄÄÇ»ÅÍ
´Ý±â

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ÇÐȸÁö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ÇÐȸÁö > µ¥ÀÌÅͺ£À̽º ¿¬±¸È¸Áö(SIGDB)

µ¥ÀÌÅͺ£À̽º ¿¬±¸È¸Áö(SIGDB)

Current Result Document :

ÇѱÛÁ¦¸ñ(Korean Title) GPU ±â¹Ý °í¼º´É ´ë·® À̵¿°´Ã¼ ¾Ïȣȭ ¾Ë°í¸®Áò
¿µ¹®Á¦¸ñ(English Title) A GPU-based High-Performance Algorithm for Encrypting a Large Number of Moving Objects
ÀúÀÚ(Author) ÃÖ¿µÈ¯   ¿©ÀºÁö   ÀÓÈ¿»ó   Young-Hwan Choi   Eunji Yeo   Hyo-Sang Lim  
¿ø¹®¼ö·Ïó(Citation) VOL 32 NO. 02 PP. 0130 ~ 0146 (2016. 08)
Çѱ۳»¿ë
(Korean Abstract)
GPU(Graphical Processing Unit)´Â ±×·¡ÇÈ Ã³¸® ÀÛ¾÷À» Àü¹®À¸·Î ´ã´çÇÏ´Â ¿¬»êÀåÄ¡·Î, ±×·¡ÇÈÀÇ Çʼö ¿ä¼ÒÀÎ ºÎµ¿¼Ò¼öÁ¡ ¿¬»êÀ» º´·ÄÀûÀ¸·Î ó¸®ÇѴٴ Ư¡À» °¡Áø´Ù. ÃÖ±Ù µé¾î GPU´Â ±× ³ôÀº º´·Äó¸® ¼º´ÉÀ» È°¿ëÇÏ¿© ±×·¡ÇÈ ÀÀ¿ë»Ó¸¸ ¾Æ´Ï¶ó ´Ù¾çÇÑ ÀÀ¿ë¿¡¼­ ó¸® ¼º´ÉÀ» ³ôÀ̱â À§ÇÏ¿© ³Î¸® È°¿ëµÇ±â ½ÃÀÛÇÏ¿´´Ù. º» ³í¹®¿¡¼­´Â ´Ù¾çÇÑ ÀÀ¿ë Áß¿¡¼­ À̵¿°´Ã¼¿¡ ´ëÇÑ ¿¬±¸¿¡ ÃÊÁ¡À» ¸ÂÃç, ½Ç½Ã°£À¸·Î »ý¼ºµÇ´Â ´ë·®ÀÇ À̵¿°´Ã¼ À§Ä¡ Á¤º¸¸¦ ¾ÏȣȭÇÏ´Â °ÍÀÇ ¼º´ÉÀ» ³ôÀ̱â À§ÇØ GPUÀÇ º´·Ä¼ºÀ» ÀÌ¿ëÇÏ´Â ¾Ë°í¸®ÁòÀ» Á¦½ÃÇÑ´Ù. ¸ÕÀú, 1,024 bit Á¤¼ö¸¦ Ç¥ÇöÇϱâ À§Çؼ­ Å« Á¤¼ö µ¥ÀÌÅÍ Ç¥Çö¸ðµ¨À» Á¦¾ÈÇÏ°í, ´ÙÀ½À¸·Î À̸¦ »ç¿ëÇÏ¿© ¾Ïȣȭ ¾Ë°í¸®ÁòÀÇ ÇÙ½É ¿¬»êÀÎ °ö¼À°ú ¸ðµâ·¯¸¦ ¼öÇàÇÏ´Â ¹æ¹ýÀ» Á¦½ÃÇÑ´Ù. Á¦¾ÈÇÏ´Â ¾Ë°í¸®ÁòÀº º´·ÄÈ­ ¼öÁØÀ» µÎ °¡Áö·Î ±¸ºÐÇÏ¿© 1) °´Ã¼ ¼öÁØ º´·ÄÈ­(object level parallelism) ¹æ¹ý°ú 2) ¿¬»ê ¼öÁØ º´·ÄÈ­(operational level parallelism) ¹æ¹ýÀ» °¢°¢ Á¦½ÃÇÑ´Ù. ±×¸®°í ½ÇÇèÀ» ÅëÇØ °¢ ¹æ¹ýÀÇ Æ¯Â¡°ú Àå´ÜÁ¡À» ºÐ¼®ÇÏ°í ±âÁ¸ÀÇ CPU±â¹Ý ¾Ë°í¸®Áò°úÀÇ ¼º´É Â÷À̸¦ º¸¿© ÇØ´ç ¾Ë°í¸®ÁòÀÇ À¯¿ë¼ºÀ» º¸ÀδÙ.
¿µ¹®³»¿ë
(English Abstract)
GPU(Graphical Processing Unit) is a special purpose electronic circuit designed to efficiently process graphical operations such as floating point operations. The efficiency of GPU comes from the high-parallelism of operations and high-bandwidth of memory communications. Using these advantages of GPUs, recently GPU is widely utilized not only in graphic applications but also in many other applications so as to increase processing performance. In this paper, we focus on moving object applications and use the GPUs¡¯ parallelism in order to increase the performance of encrypting a large number of moving objects generated in real time. First, we propose a data model for representing a large(1,024 bit) integer. Next, based on the data model, we provide algorithms for multiplication and modular operations which are essential in the data encryption. We then provide two kinds of parallel algorithms for these operations: 1) object level parallelism algorithm and 2) operational level parallelism. Finally, we provide experiment results to show the characteristics and the performance benefits of the proposed algorithms.
Å°¿öµå(Keyword) GPU   À̵¿°´Ã¼   RSA ¾Ïȣȭ ¾Ë°í¸®Áò   º´·Ä󸮠  GPU   Moving Objects   RSA Cryptograph Algorithm   Parallel Processing  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå