Ç¥ÁØÈ­ Âü¿©¾È³»

TTAÀÇ Ç¥ÁØÇöȲ

Ȩ > Ç¥ÁØÈ­ °³¿ä > TTAÀÇ Ç¥ÁØÇöȲ

Ç¥ÁعøÈ£ TTAK.KO-10.1309 ±¸Ç¥ÁعøÈ£
Á¦°³Á¤ÀÏ 2021-12-08 ÃÑÆäÀÌÁö 6
ÇѱÛÇ¥ÁØ¸í °¡º¯ÆÞ½ºÆø º¯Á¶±â¹Ý ±¤ÇÐ Ä«¸Þ¶ó Åë½ÅÀ» À§ÇÑ ½Å°æ¸Á ¾ÆÅ°ÅØó
¿µ¹®Ç¥Áظí Neural Network Architecture for Variable Pulse Width Modulation based Optical Camera Communication
Çѱ۳»¿ë¿ä¾à ÀÌ Ç¥ÁØ¿¡¼­´Â Á¤Àû ¹× À̵¿ Á¶°Ç ¸ðµÎ¿¡¼­ LED¸¦ Á¤È®ÇÏ°Ô °ËÃâÇÒ ¼ö ÀÖ´Â ¼ö½Å±â ½Å°æ¸Á(NN)À» °³¹ßÇÑ´Ù. ±âÁ¸¿¡´Â LED ¾î·¹À̸¦ ´ÜÀÏ LED·Î °¨ÁöÇßÁö¸¸, ¿©±â¼­´Â LED ¾î·¹ÀÌÀÇ °¢ LED¸¦ Á¦¾ÈµÈ NN ±â¹ýÀ» »ç¿ëÇÏ¿© °ËÃâÇÒ ¼ö ÀÖ´Ù. NNÀº ¿©·¯ °ø°£ °èÃþ¿¡¼­ ÁÖ¸ñÇÒ ¸¸ÇÑ Æ¯¼ºÀ» °¡Áø ´ë±Ô¸ð À̹ÌÁö µ¥ÀÌÅÍ ¼ÂÀ¸·Î ºÐ·ùÇÏ°í ÈĹæ ÀüÆĸ¦ ÅëÇØ µ¥ÀÌÅ͸¦ ÀÚµ¿À¸·Î ÇнÀÇÒ ¼ö ÀÖ´Ù. ±âÁ¸¿¡´Â LED ¾î·¹À̸¦ °¨ÁöÇÏ°í À̵¿¼ºÀ» Áö¿øÇϱâ À§ÇÑ ´Ù¾çÇÑ NNÀÌ Á¦½ÃµÇ¾î ÀÖ´Ù. ´Ù¸¸ ƯÁ¤ LED¸¦ Â÷´ÜÇÑ Ã¤ ¼Õ°¡¶ôÀ» °è¼Ó ¿òÁ÷¿© ¸ð¹ÙÀÏ ½Ã³ª¸®¿À¸¦ Á¦¾ÈÇÑ´Ù. ÀÌ·¯ÇÑ ¹æ¹ýÀ» »ç¿ëÇÏ¸é ¿À·ùÀ²ÀÌ ³ôÁö¸¸ º» Ç¥ÁØ¿¡¼­ Á¦¾ÈÇÏ´Â ¹æ½ÄÀº Ä«¸Þ¶ó ÀÚü¸¦ ¿òÁ÷¿© ¸ð¹ÙÀÏ ½Ã³ª¸®¿À¸¦ Àû¿ëÇØ ¶Ù¾î³­ ºñÆ®¿À·ù·ü(BER)À» ´Þ¼ºÇß´Ù.
¿µ¹®³»¿ë¿ä¾à In this standard, we develop a neural network (NN) at the receiver for accurately detecting the LED in both static and mobile conditions. In existing literature, an LED array is detected as a single LED, however, each LED of the LED array can be detected using the proposed NN technique. The NN can classify large image datasets with remarkable characteristics in several spatial layers and automatically learn from data through back propagation. Different NNs have been presented in the literatures for the detection of the LED array and to support mobility. However, they simulated the mobile scenario by moving a finger continuously, blocking the particular LEDs. The error rate was high using those methods. In our proposed scheme, we have simulated the mobile scenario by moving the camera itself and achieved an excellent BER.
±¹Á¦Ç¥ÁØ
°ü·ÃÆÄÀÏ TTAK.KO-10.1309.pdf TTAK.KO-10.1309.pdf            

ÀÌÀü
i-PIN ¼­ºñ½º Àü´Þ ¸Þ½ÃÁö Çü½Ä
´ÙÀ½
ÄÄÇ»ÅÍ Æ÷·»½ÄÀ» À§ÇÑ µðÁöÅÐ µ¥ÀÌÅÍ ¼öÁýµµ±¸ ¿ä±¸»çÇ×