A Novel Rapid-Flooding Approach With Real-Time Delay Compensation for Wireless-Sensor Network Time Synchronization

Fanrong Shi, Simon X. Yang, Xianguo Tuo, Lili Ran, Yuqing Huang

Abstract: One-way-broadcast-based flooding time synchronization algorithms are commonly used in wireless-sensor networks (WSNs). However, the packet delay and clock drift pose a challenge to accuracy, as they entail serious by-hop error accumulation problems in the WSNs. To overcome this, a rapid-flooding multibroadcast time synchronization with real-time delay compensation (RDC-RMTS) is proposed in this article. By using a rapid-flooding protocol, flooding latency of the referenced time information is significantly reduced in the RDC-RMTS. In addition, a new joint clock skew-offset maximum-likelihood estimation (MLE) is developed to obtain the accurate clock parameter estimations and the real-time packet delay estimation. Moreover, an innovative implementation of the RDC-RMTS is designed with an adaptive clock offset estimation. The experimental results indicate that the RDC-RMTS can easily reduce the variable delay and significantly slow the growth of by-hop error accumulation. Thus, the proposed RDC-RMTS can achieve accurate time synchronization in large-scale complex WSNs.

The Challenge of Multi-Hop Synchronization

In Wireless Sensor Networks (WSNs), nodes require a shared sense of time for tasks like coordinated data sensing or industrial control. "Flooding" is a common synchronization method where a root node broadcasts its time, and other nodes pass it along.

However, two main issues degrade accuracy as the signal travels further from the root:

  1. Clock Drift: Every node’s crystal oscillator runs at a slightly different speed (skew).
  2. Packet Delay: The time it takes to process and transmit a packet varies.

Existing state-of-the-art methods like PulseSync or RMTS use "rapid flooding" to minimize wait times, but they usually rely on a fixed, pre-calibrated delay constant ($D_{fixed}$). If the hardware changes or the environment becomes noisy, this fixed constant becomes inaccurate, causing errors to pile up at every hop—the "by-hop error accumulation" problem.

Proposed Method: RDC-RMTS

The authors introduce RDC-RMTS (Rapid-flooding multi-broadcast time synchronization with real-time delay compensation) to move beyond fixed delay constants. The system constructs a "two-way message exchange" model using only one-way broadcasts. By observing multiple redundant broadcasts from neighbors, a node can mathematically solve for the transmission delay on the fly.

The RDC-RMTS synchronization model, showing the multi-broadcast MLE for clock skew and the two-way message exchange for offset estimation.

Joint Clock Skew-Offset MLE

A key innovation is the new joint Maximum Likelihood Estimation (MLE) algorithm. Unlike linear regression (used in FTSP) which can be slow to converge, this joint MLE estimates both the clock's speed (skew) and its time difference (offset) simultaneously and independently. This independence avoids "hunting" behavior where an error in one parameter causes the other to oscillate, leading to much faster convergence.

Performance and Error Accumulation

A critical advantage of the proposed method is its handling of error growth across hops. Theoretical comparisons show that while traditional protocols like FTSP and PulseSync have errors related to total delay $D$, the RMTS approach reduces this to a much smaller variable component $d$, significantly slowing the rate of error accumulation as the network size increases.

Comparison of synchronization error across single-hop and k-hop scenarios for various protocols.

Key Results

The researchers evaluated RDC-RMTS using a physical testbed of 25 nodes in a 24-hop line topology and large-scale simulations with 300 nodes.

  • Superior Accuracy: In the 24-hop network, RDC-RMTS maintained a mean global synchronization error of approximately 8 µs. For comparison, the classic FTSP algorithm had an error of 83 µs, and PulseSync was around 15 µs.
  • Rapid Convergence: RDC-RMTS reaches a stable state in just 2 to 7 synchronization rounds, whereas older methods like FTSP often require more than 12 rounds.
  • Scalability: The "error growth curve" for RDC-RMTS is significantly flatter than its competitors, making it much better suited for very large industrial plants or structural health monitoring.

Conclusion

RDC-RMTS provides a "plug-and-play" synchronization solution that adapts to hardware and environmental changes without manual calibration. While the multi-broadcast nature uses more radio energy, the gains in accuracy and convergence speed make it a powerful tool for large-scale complex WSNs.

Summary updated

Abstract

One-way-broadcast-based flooding time synchronization algorithms are commonly used in wireless-sensor networks (WSNs). However, the packet delay and clock drift pose a challenge to accuracy, as they entail serious by-hop error accumulation problems in the WSNs. To overcome this, a rapid-flooding multibroadcast time synchronization with real-time delay compensation (RDC-RMTS) is proposed in this article. By using a rapid-flooding protocol, flooding latency of the referenced time information is significantly reduced in the RDC-RMTS. In addition, a new joint clock skew-offset maximum-likelihood estimation (MLE) is developed to obtain the accurate clock parameter estimations and the real-time packet delay estimation. Moreover, an innovative implementation of the RDC-RMTS is designed with an adaptive clock offset estimation. The experimental results indicate that the RDC-RMTS can easily reduce the variable delay and significantly slow the growth of by-hop error accumulation. Thus, the proposed RDC-RMTS can achieve accurate time synchronization in large-scale complex WSNs.
BibTeX
@article{shi2020novel,
  author = {Fanrong Shi and Simon X. Yang and Xianguo Tuo and Lili Ran and Yuqing Huang},
  title = {{A Novel Rapid-Flooding Approach With Real-Time Delay Compensation for Wireless-Sensor Network Time Synchronization}},
  journal = {IEEE Transactions on Cybernetics},
  year = {2020},
  doi = {10.1109/tcyb.2020.2987758},
}

Loading related content...

Figure