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

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ³í¹®Áö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ³í¹®Áö > Çѱ¹Á¤º¸°úÇÐȸ ³í¹®Áö > Á¤º¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Á¤º¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Current Result Document : 9 / 61 ÀÌÀü°Ç ÀÌÀü°Ç   ´ÙÀ½°Ç ´ÙÀ½°Ç

ÇѱÛÁ¦¸ñ(Korean Title) IoT µð¹ÙÀ̽º¿¡¼­ ¾²·¹µå µ¿ÀÛÀ» ÃÖÀûÈ­Çϱâ À§ÇÑ CPU ÁßÀçÀÚ
¿µ¹®Á¦¸ñ(English Title) CPU Mediator for Optimizing Thread Operations on IoT Devices
ÀúÀÚ(Author) ÀӱٽĠ  °­µ¿Çö   ¾ö¿µÀÍ   Geunsik Lim   Dong Hyun Kang   Young Ik Eom  
¿ø¹®¼ö·Ïó(Citation) VOL 25 NO. 12 PP. 0616 ~ 0621 (2019. 12)
Çѱ۳»¿ë
(Korean Abstract)
¿î¿µÃ¼Á¦ ±â¼úÀº ÀüÅëÀûÀ¸·Î °í¼º´É ÄÄÇ»Æà ȯ°æÀ» ±â¹ÝÀ¸·Î ¹ßÀüÇØ ¿ÔÀ¸¸ç, ÃÖ±ÙÀÇ ¿î¿µÃ¼Á¦µéÀº ¿î¿µÃ¼Á¦ ³»ºÎÀÇ ¾²·¹µå °£ ¼º´É °£¼·À» ÁÙÀ̱â À§ÇØ ÁÖ·Î 1:1 ¾²·¹µå ¸ðµ¨À» äÅÃÇÏ°í ÀÖ´Ù. ±×·¯³ª, ±âÁ¸ ¾²·¹µå ¸ðµ¨Àº Á¦ÇÑµÈ Çϵå¿þ¾î ÀÚ¿øÀ» °¡Áø IoT µð¹ÙÀ̽º ȯ°æ¿¡¼­´Â ½Ã½ºÅÛÀÇ ¼º´ÉÀ» ÀúÇϽÃÅ°´Â ¹®Á¦Á¡À» °¡Áö°í ÀÖ´Ù. ÀÌ¿¡, º» ³í¹®¿¡¼­´Â °í¼º´É ±â¹Ý 1:1 ¾²·¹µå ¸ðµ¨À» ºÐ¼®ÇÏ°í À̸¦ ±â¹ÝÀ¸·Î IoT µð¹ÙÀ̽º ȯ°æ¿¡ ÀûÇÕÇÑ ¾²·¹µå µ¿ÀÛ ÃÖÀûÈ­ ±â¹ýÀ» Á¦¾ÈÇÑ´Ù. ƯÈ÷, Á¦¾È ±â¹ýÀº °³¹ßÀÚ¿¡°Ô POSIX ȣȯ¼ºÀ» °®´Â »ç¿ëÀÚ ·¹º§ÀÇ ¾²·¹µå ½ºÄÉÁÙ¸µ API¸¦ Á¦°øÇϱ⠶§¹®¿¡ °³¹ßÀÚµéÀº ´ÜÀÏÈ­µÈ ÃÖÀûÈ­ ÀÎÅÍÆäÀ̽º¸¦ ÅëÇØ ´Ù¾çÇÑ IoT µð¹ÙÀ̽ºµéÀÇ ¾²·¹µå ¼º´ÉÀ» ÀÏ°üµÇ°Ô °³¼± ¹× Á¦¾îÇÒ ¼ö ÀÖ´Ù. º» ³í¹®¿¡¼­ Á¦¾ÈÇÏ´Â ±â¹ýÀº Raspberry Pi3 ȯ°æ¿¡¼­ ±¸ÇöµÇ¾úÀ¸¸ç, ½ÇÇè °á°ú CPU ºÎÇÏ°¡ ³ôÀº ÀÛ¾÷ ȯ°æ¿¡¼­ ¾²·¹µåÀÇ ÀÀ´ä ½Ã°£À» ±âÁ¸ ¾²·¹µå ¸ðµ¨ ´ëºñ 7.3¹è ÁÙÀÏ ¼ö ÀÖÀ½À» È®ÀÎÇÏ¿´´Ù.
¿µ¹®³»¿ë
(English Abstract)
The technologies of the operating systems have traditionally been developed based on the high-performance computing environments and they have been adopting the 1:1 thread mapping model to mitigate interference among the threads in the operating system. However, the conventional thread model reveals performance problems in the IoT environments because of its limited hardware resources. In this paper, we first study the 1:1 thread mapping model based on the high-performance computing systems and then propose a new scheme that optimizes the behavior of thread management on IoT devices. Particularly, since the proposed technique provides developers with a POSIX-compatible user-level thread scheduling API, the developers can control and improve the thread performance of various IoT devices with the unified optimization interface. We implemented the proposed scheme on Raspberry Pi3 and clearly confirmed through experiments that our scheme can reduce the response time of the threads 7.3 times on average, compared with the conventional thread model under the CPU-intensive workloads.
Å°¿öµå(Keyword) ¾²·¹µå ¸ðµ¨   ¾²·¹µå ÃÖÀûÈ­   ¾²·¹µå ÇÁ·ÎÆÄÀϸµ   ¾²·¹µå ½ºÄÉÁÙ¸µ   thread model   thread optimization   thread profiling   thread scheduling  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå