Towards Trajectory Prediction-based UAV Deployment in Smart Transportation Systems
| dc.contributor.author | Liang, Fan | |
| dc.date.accessioned | 2023-11-20T20:27:49Z | |
| dc.date.available | 2023-11-20T20:27:49Z | |
| dc.date.issued | 2023-10-15 | |
| dc.description.abstract | A smart transportation system (i.e., intelligent transportation system) refers to a transportation critical infrastructure system that integrates advanced technologies (e.g., networking, distributed computing, big data analytics, etc.) to improve the efficiency, safety, and sustainability of the transportation system. However, the rapid increase in the number of vehicles on roads and significant fluctuations in the flow of traffic can cause the coverage holes of Road Side Units (RSUs) and local traffic overload in smart transportation systems, which can negatively affect the performance of systems and causes accidents. To address these issues, deploying Unmanned Aerial Vehicles (UAVs) as mobile RSUs is a viable approach. Nonetheless, how to deploy UAVs to the optimal position in the smart transportation system remains an unsolved issue. This paper proposes a Vehicle Trajectory-based Dynamic UAV Deployment Algorithm (VTUDA). The VTUDA utilizes vehicle trajectory prediction information to improve the efficiency of UAV deployment. First, we deploy a distributed Seq2Seq-GRU model to the UAVs and train the model. We leverage the well-trained model to predict vehicle trajectory. VTUDA then uses the predicted information to make informed decisions on the optimal location to position the UAVs. Further- more, VTUDA considers both the condition of communication channels and energy consumption during the deployment process to ensure that UAVs are deployed to optimal positions. Our experimental results confirm that the proposed VTUDA can effectively improve the deployment of UAVs. The experimental results also demonstrate that VTUDA can significantly enhance vehicle access and communication quality between vehicles and UAVs. | |
| dc.description.provenance | Made available in DSpace on 2023-11-20T20:27:49Z (GMT). No. of bitstreams: 1 Towards Trajectory Prediction.pdf: 947125 bytes, checksum: e846db905466187b2f22f732bfc57522 (MD5) Previous issue date: 2023-10-15 | en |
| dc.identifier.citation | Liang, F., Liu, X., Kose, N. A., Gundogan, K., Yu, W. (2023) Towards Trajectory Prediction-based UAV Deployment in Smart Transportation Systems. (Report No. IHS/CR-2023-1010). The Sam Houston State University Institute for Homeland Security. | |
| dc.identifier.uri | https://hdl.handle.net/20.500.11875/4245 | |
| dc.language.iso | en_US | |
| dc.publisher | Institute for Homeland Security | |
| dc.relation.ispartofseries | IHS; CR-2023-1010 | |
| dc.subject | Smart Transportation Systems, Edge Comput- ing, UAV Deployment, Machine Learning | |
| dc.title | Towards Trajectory Prediction-based UAV Deployment in Smart Transportation Systems | |
| dc.type | Article |