Calibrating adaptive feature compression for edge vehicle perception under 8% bandwidth loss: a sensitivity sweep
PDF

Keywords

edge vehicle perception
adaptive feature compression
bandwidth loss
paired simulation
reproducibility

Abstract

We evaluated adaptive feature compression for edge vehicle perception under 8% bandwidth loss. A deterministic paired simulation generated 56 cases and preserved a upper-severity quartile. Mean latency-adjusted recall changed from 0.572 to 0.626; the paired difference was +0.055 (95% interval +0.052 to +0.057). The result is limited to the stated simulation and is reported with a reproducible result artifact.

PDF

References

Liu, M., Que, H., Fan, D., Gao, H., Zhu, T., Yao, H., Zhang, Q., Zhang, R., Huang, X., & Qiao, F. (2026). ACEsplat: Accelerated 3D Gaussian Scene Regression via RGB and Poses Only. arXiv. https://doi.org/10.48550/arXiv.2606.22091

Mansour, I., & Singh, S. (2025). Role of AI in an Autonomous Vehicle Perception Systems: Use of Convolutional Neural Network Approach to Design the Vehicle Perception System. SAE Technical Paper Series, 1, 2025-01-8012. https://doi.org/10.4271/2025-01-8012

Sun, Z., Zha, L., & Zhou, Y. (2023). Application of Edge Intelligence and Vehicle-To-Vehicle Communication in Next-Generation Autonomous Driving. Highlights in Science, Engineering and Technology, 38, 708-715. https://doi.org/10.54097/hset.v38i.5935

Cao, L. (2025). Attention Based Tracking Head for Multiple Object Tracking in Autonomous Vehicle Perception Systems. https://doi.org/10.36227/techrxiv.174198467.74129950/v1

Zhou, L. (2023). A Vehicle Service Migration Strategy Algorithm in 5G NR-V2X. Frontiers in Computing and Intelligent Systems, 2(3), 48-55. https://doi.org/10.54097/fcis.v2i3.5211

Tang, B., Yu, Q., & Liu, Z. (2026). Drivers’ Safety Perception in Autonomous Vehicle Road Sharing: A Knowledge-Segmented TPB and Ordered Logit Analysis. Applied Sciences, 16(7), 3599. https://doi.org/10.3390/app16073599

ВЛАДЫКО, МУТХАННА, & КУЧЕРЯВЫЙ (2020). TRAFFIC OFFLOADING METHOD FOR V2X/5G NETWORKS BASED ON THE EDGE COMPUTING SYSTEM. Электросвязь(8(9)). https://doi.org/10.34832/elsv.2020.9.8.004

Chang, R. I., Hsu, T. W., Yang, C., & Chen, Y. T. (2025). Bounded-Error LiDAR Compression for Bandwidth-Efficient Cloud-Edge In-Vehicle Data Transmission. Electronics, 14(5), 908. https://doi.org/10.3390/electronics14050908

Miucic, R., & Rajab, S. (2022). Computer Vision-Based V2X Collaborative Perception. SAE Technical Paper Series, 1, 2022-01-0073. https://doi.org/10.4271/2022-01-0073

Liu, C., Zuo, H., Yao, J., Li, Y., & Jiang, F. (2025). Attention-aware upsampling-downsampling network for autonomous vehicle vision-based multitask perception. Complex & Intelligent Systems, 11(6), 279. https://doi.org/10.1007/s40747-025-01870-4