To satisfactorily gratify the scope of International Mobile Telecommunications 2020 (IMT-2020), 3GPP has launched the standardization activity of the fifth generation (5G) New Radio (NR) to deploy the first phase system (Release 15) in 2018 and the ready system (Release 16) in 2020. Different from the IMT-Advanced system solely enhancing the transmission data rates regardless the variety of emerging wireless traffic, the IMT-2020 system supports diverse wireless services including enhanced mobile broadband (eMBB), massive machine-type communications (mMTC) and ultra-reliable and low latency communications (URLCC) to fully capture wireless applications in 2020. Among all the wireless services, URLLC jointly demanding low latency and high reliability and mMTC emphasizing on high reliability create substantial impacts to the designs of NR air interface. On the advert of the conventional feedback based transmission in LTE/LTE-A designed for eMBB imposing potential inefficiency in supporting URLLC, in this paper, we revisit the feedbackless transmission framework, and reveal a tradeoff between these two transmission frameworks to latency and reliability guarantees. A multi-armed bandit (MAB) based reinforcement learning approach is therefore proposed to achieve the optimum harmonization of feedback and feedbackless transmissions. Our simulation results fully demonstrate the practicability of the proposed approach in supporting URLLC, to justify the potential of our approach in the practice of 5G NR.
All Science Journal Classification (ASJC) codes
- Information Systems
- Hardware and Architecture
- Computer Networks and Communications