车车通信作为我国下一代高速铁路5G-R无线通信系统的关键技术，实现车车通信的时间同步对列车行车安全至关重要。针对车车通信时间同步易受非平稳无线信道和报文传输时延的影响，导致车车通信时间同步性能低的问题，提出了一种变分模态分解与时间注意力双向长短期记忆网络(TA-BLSTM)增强的车车通信时间同步方法。首先，通过对车车通信时延误差进行分析，建立了5G-R车车通信时钟模型。其次，采用变分模态分解模型将车车通信时间序列分解为不同频率的本征模态，从而分离出噪声成分，提高信号的信噪比。然后，通过计算能量值识别出噪声主导分量，利用小波软阈值降噪法对噪声分量进行去噪，提高车车通信同步时间序列质量。最后，提出一种融合时间注意力机制的双向长短期记忆TA-BLSTM网络，通过双向LSTM网络提取车车通信时间同步序列的长期时域特征，并采用时间注意力机制动态捕获时间同步序列的时序依赖特性，实现对车车通信时间同步偏差的高精度预测与动态补偿，从而完成车车通信时间同步。通过仿真实验，所提方法在有中继和无中继场景下均能有效实现车车时间同步。与其他方法相比，所提方法能够有效降低同步偏移误差，在车车通信时间同步过程中具有更快的收敛速度和稳定性。

Train-to-train communication constitutes the foundational architecture for China’s railway-dedicated 5G-R communication systems, and achieving time synchronization for train to train communication is crucial for train operation safety. To address the issue of poor time synchronization performance caused by non-stationary wireless channels and transmission delays in train to train communication, this paper proposes an enhanced time synchronization approach based on variational mode decomposition combined with a bidirectional long short-term memory network incorporating time attention mechanisms. First, a 5G-R train-to-train communication clock model is established by analyzing the delay errors in train-to-train communication. Then, the VMD model is employed to decompose the train-to-train communication time series into intrinsic mode functions of different frequencies, thus isolating noise elements and enhancing the signal-to-noise ratio. Next, noise-dominated components are identified by calculating energy values, and wavelet soft thresholding is applied to denoise these components, enhancing the quality of the train to train communication synchronization time series. Finally, a TA-BLSTM network is proposed, which integrates a time attention mechanism into a bidirectional LSTM framework. This network extracts long-term temporal features from the train to train time synchronization sequence using the bidirectional LSTM, while the time attention mechanism dynamically captures temporal dependencies, enabling high-precision prediction and dynamic compensation of time synchronization deviations in train to train communication, thus achieving accurate time synchronization. Simulation experiments demonstrate that the proposed method can effectively achieve train-to-train time synchronization in both relay and non-relay communication scenarios. Compared with other methods, the proposed approach significantly reduces synchronization offset errors and offers faster convergence speed and greater stability during the train to train time synchronization process.