Loading web-font TeX/Math/Italic
Joint Blockage Prediction and UAV Deployment in Millimeter-Wave Communication Networks | IEEE Journals & Magazine | IEEE Xplore

Joint Blockage Prediction and UAV Deployment in Millimeter-Wave Communication Networks


Abstract:

Leveraging the flexibility of unmanned aerial vehicles (UAV) and the large bandwidth of millimeter-wave (mmWave), UAV mmWave communications can provide efficient data tra...Show More

Abstract:

Leveraging the flexibility of unmanned aerial vehicles (UAV) and the large bandwidth of millimeter-wave (mmWave), UAV mmWave communications can provide efficient data transmission in various on-demand services or emergency situations. However, mmWave signals are sensitive to blockages. Existing solutions to reduce blockage are to deploy UAVs using a probabilistic blockage model, which may fail to match the actual blockage status. To address this limitation, this paper proposes a practical scheme to predict the link blockage and deploy multiple UAVs to provide anti-blockage mmWave transmission for terrestrial users (UEs). It can accurately predict link blockage via location information of UAVs and UEs, along with historical blockage status without prior knowledge of building blockages, such as shape, size, or location. Particularly, the scheme schedules UAVs to fly over the target area, gathering the positions of UEs and blockage status, which is then employed to train a link geometric feature based blockage prediction model to acquire the blockage status for any desired UAV locations. Then, a multi-agent Q-learning based method is adopted to determine the positions of UAVs through an iterative process that alternates between position optimization and blockage prediction. Simulation results validate that the proposed scheme can effectively avoid blockages and enhance the sum rate compared to benchmark methods.
Published in: IEEE Transactions on Vehicular Technology ( Volume: 73, Issue: 11, November 2024)
Page(s): 17881 - 17886
Date of Publication: 18 July 2024

ISSN Information:

Funding Agency:


References

References is not available for this document.