Abstract
Vehicle-to-Everything (V2X) systems that support cooperative perception and control require precise network-wide time synchronization. In dense traffic, however, existing schemes often decouple synchronization from communication delay. As a result, rapidly varying uplink queueing delays remain uncompensated and introduce persistent asymmetric bias that degrades synchronization accuracy. To address this issue, we have proposed Queue-Sync, a delay-compensated time synchronization framework for a Cloud-RSU-EV (CRE) architecture. Queue-Sync consists of a roadside Dynamic Delay Prediction Module (DPM) and an on-vehicle Delay-Compensated Synchronization Module (DCM). The DPM combines an M/M/1 queueing model with an adaptive extended Kalman filter (AEKF) to estimate the uplink arrival rate and queueing delay online. The DCM performs asymmetric timestamp-based synchronization. It injects the predicted delay into the offset calculation for feedforward compensation, and feeds back the measured round-trip time (RTT) to calibrate the DPM. Based on an extended fine-grained clock model, simulation results show that Queue-Sync consistently yields a lower average synchronization error compared to baseline schemes. This improvement holds under different synchronization intervals and traffic densities. It achieves a smoother convergence and more stable clock synchronization performance at the cost of a modest increase in synchronization message overhead.
| Original language | English |
|---|---|
| Pages (from-to) | 969-974 |
| Number of pages | 6 |
| Journal | Proceedings of the International Conference on Computer Supported Cooperative Work in Design, CSCWD |
| Issue number | 2026 |
| Early online date | 6 Jul 2026 |
| DOIs | |
| Publication status | Published - 6 Jul 2026 |
| Event | 29th International Conference on Computer Supported Cooperative Work in Design, CSCWD 2026 - Fuzhou, China Duration: 13 May 2026 → 15 May 2026 |
Keywords
- delay compensation
- extended Kalman filter
- queueing delay
- time synchronization
- Vehicle-to-Everything (V2X)
Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver