2019³â ÄÄÇ»ÅÍÁ¾ÇÕÇмú´ëȸ
ÇѱÛÁ¦¸ñ(Korean Title) |
Dynamic Resource Slicing of eMBB/URLLC Traffics in 5G Wireless Networks: A Reinforcement Learning Based Approach |
¿µ¹®Á¦¸ñ(English Title) |
Dynamic Resource Slicing of eMBB/URLLC Traffics in 5G Wireless Networks: A Reinforcement Learning Based Approach |
ÀúÀÚ(Author) |
Madyan Alsenwi
Shashi Raj Pandey
Yan Kyaw Tun
Choong Seon Hong
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 46 NO. 01 PP. 1233 ~ 1235 (2019. 06) |
Çѱ۳»¿ë (Korean Abstract) |
Ultra Reliable Low Latency Communications (URLLC) is a 5G wireless networks service that accommodates applications having strict reliability and latency requirements such as industrial IoT, virtual reality, and autonomous vehicles. URLLC aims low latency and high reliability transmissions. In this paper, the dynamic multiplexing of URLLC and enhanced Mobile Broad Band (eMBB) is tackled. We model the problem as a Markov Decision Process. Moreover, We propose a weighted formulation that takes into account both the total average data rate of eMBB traffic and the URLLC reliability. To this end, We propose a Q-learning algorithm where the gNB learns the optimal resource allocation to eMBB and URLLC traffics. The simulation results illustrate the performance of the proposed algorithm. |
¿µ¹®³»¿ë (English Abstract) |
|
Å°¿öµå(Keyword) |
5G New Radio
eMBB
URLLC
Resource Slicing
Q-learning
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|