A Feasible Study of Cube Sensing Organization Map for Cognitive Spectrum Allocation
- Authors
- Pham, Quang-Dung; Do, Trong-Hop; Dao, Nhu-Ngoc; Na, Woongsoo; Lee, Chunghyun; Cho, Sungrae
- Issue Date
- Jan-2021
- Publisher
- IEEE
- Keywords
- Cognitive spectrum sensing; probabilistic sensing; sensing organization map; reinforcement learning
- Citation
- 35TH INTERNATIONAL CONFERENCE ON INFORMATION NETWORKING (ICOIN 2021), pp 858 - 863
- Pages
- 6
- Journal Title
- 35TH INTERNATIONAL CONFERENCE ON INFORMATION NETWORKING (ICOIN 2021)
- Start Page
- 858
- End Page
- 863
- URI
- https://scholarworks.bwise.kr/cau/handle/2019.sw.cau/47522
- DOI
- 10.1109/ICOIN50884.2021.9333998
- Abstract
- Spectrum availability has been considered as one of the most crucial factors for the success of serving dense user devices in fifth generation (5G) mobile communications and beyond. To this end, cognitive spectrum sensing technologies can be utilized to detect unused wireless channels in the coverage area of the user devices, which are possibly used for device-to-device communications efficiently. However, frequent wide range sensing of spectrum extremely exhausts energy resource of the user devices. In this paper, we propose a probabilistic cognitive spectrum sensing (PCSS) mechanism to overcome this challenge. In the PCSS mechanism, a cube sensing organization map (CSOM) is dimensionalized on the basis of channels and locations in the coverage area. Weight of each block in the CSOM represents the probability that a user device successfully senses the corresponding channel idle in the corresponding location. The weight of each block is learned after every sensing periods to reinforce the CSOM. Consequently, user devices sense the list of channels in their proximity with the probabilities indicated by the weight of the best matching CSOM blocks. Simulation results show that the PCSS mechanism overcomes the existing techniques in terms of energy efficiency and positive sensing ratio.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - College of Software > School of Computer Science and Engineering > 1. Journal Articles
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.