9th International Conference on Cognitive Radio Oriented Wireless Networks

Research Article

Indoor Propagation Model for TV White Space

Download606 downloads
  • @INPROCEEDINGS{10.4108/icst.crowncom.2014.255380,
        author={Wataru Yamada and Motoharu Sasaki and Takatoshi Sugiyama and Oliver Holland and Shuyu Ping and Bright Yeboah-Akowuah and Jiwon Hwang and Hamid Aghvami},
        title={Indoor Propagation Model for TV White Space},
        proceedings={9th International Conference on Cognitive Radio Oriented Wireless Networks},
        keywords={frequency sharing systems tvws propagation measurement indoor propagation model interference evaluation},
  • Wataru Yamada
    Motoharu Sasaki
    Takatoshi Sugiyama
    Oliver Holland
    Shuyu Ping
    Bright Yeboah-Akowuah
    Jiwon Hwang
    Hamid Aghvami
    Year: 2014
    Indoor Propagation Model for TV White Space
    DOI: 10.4108/icst.crowncom.2014.255380
Wataru Yamada1,*, Motoharu Sasaki1, Takatoshi Sugiyama1, Oliver Holland2, Shuyu Ping2, Bright Yeboah-Akowuah2, Jiwon Hwang2, Hamid Aghvami2
  • 1: NTT Corporation
  • 2: King’s College London
*Contact email: yamada.wataru@lab.ntt.co.jp


In this paper, indoor propagation characteristics for TV white space (TVWS) frequency bands are studied. Measurements in the actual indoor environment are performed and described. An indoor propagation model for TVWS is developed, which is useful for evaluating secondary-secondary interference in TVWS scenarios, and its empirical parameters are derived according to measurement results. Our proposed model is based on components of free space path loss, penetration losses of walls and floors, an attenuation coefficient against distance, and an attenuation constant. Frequency dependency of these parameters is derived. Our proposed model is compared with representative indoor models from the literature. Results show that our proposed model, with the obtained empirical parameters, achieves better performance than representative indoor models. The standard deviation of estimation error for our proposed model is only 5.9 dB.