Paper Info
Title
Adaptive Repetition Scheme with Machine Learning for 3GPP NB-IoT
Abstract
In NB-IoT systems, UEs with poor signal quality employ more repetitions to compensate for additional signal attenuation. Excessively high CE levels and repetitions of UEs lead to wastage of valuable wireless resources, whereas inadequate CE levels and repetitions result in data retrieval failure at the receiving end. Therefore, a machine learning-based adaptive repetition scheme for a 3GPP NB-IoT system is proposed in this work to effectively improve overall network transmission efficiency. The results of simulation show the effect of the discount factor? on the convergence behavior of the proposed scheme, with a lower discount factor value denoting the myopic behavior of the proposed scheme, which results from the fact that it places more emphasis on immediate rewards. And the propose scheme is capable of effectively improving the average spectral efficiency.
Year
DOI
Venue
2018
10.1109/PRDC.2018.00046
2018 IEEE 23rd Pacific Rim International Symposium on Dependable Computing (PRDC)
Keywords
Field
DocType
Reinforcement learning,Signal to noise ratio,3GPP,Interference,Wireless communication,Convergence
Convergence (routing),Wireless,Discounting,Computer science,Data retrieval,Signal-to-noise ratio,Real-time computing,Artificial intelligence,Spectral efficiency,Interference (wave propagation),Machine learning,Reinforcement learning
Conference
ISSN
ISBN
Citations 
1555-094X
978-1-5386-5700-3
0
PageRank 
References 
Authors
0.34
0
4
Name
Order
Citations
PageRank
Li-Sheng Chen101.01
Wei-Ho Chung251768.70
Ing-Yi Chen321523.61
Sy-Yen Kuo42304245.46