ETRI-Knowledge Sharing Plaform

KOREAN
논문 검색
Type SCI
Year ~ Keyword

Detail

Conference Paper QoS guaranteed Small Cell ON/OFF Techniques Based on Deep learning
Cited - time in scopus Share share facebook twitter linkedin kakaostory
Authors
Taeyoon Park, Eunghyo Kim, Jaewan Park, Jongwon Han, Een-Kee Hong, Soojung Jung, Taegyun Noh
Issue Date
2018-10
Citation
International Conference on Information and Communication Technology Convergence (ICTC) 2018, pp.805-808
Publisher
IEEE
Language
English
Type
Conference Paper
Abstract
Recently, there is increasing interest in how to secure network capacity to support the exponential growth of mobile data traffic. Ultra-Dense-Network (UDN) is expected to solve the problem of network capacity by installing a smaller cell with high densification. But UDN has energy efficiency problems. since many cells consume a lot of energy. To solve this problem, small cell on/off is expected to be a key technology. However, small cell on/off has also challenge in guaranteed QoS of traffics. When a small cell is off, traffic of the cell is offloaded to the neighbor cell and the QoS problem occurs. In this paper, we solve the trade-off problem between QoS and energy efficiency by turning on/off small cells in consideration of QoS by deep learning which is a representative method to solve nonlinear problems. We find the optimized weighing factors of multiple criteria for small cell on/off such as traffic load, traffic QoS, number of UEs and path loss based on learning approach.
KSP Keywords
Energy efficiency, Key technology, Learning approach, Mobile data traffic, Multiple criteria, Network capacity, Nonlinear problems, Path loss, Trade-off, Traffic Load, Ultra-dense