Integrated sensing and communication (ISAC) is a promising paradigm for next-generation vehicular networks, yet existing orthogonal frequency-division multiplexing (OFDM)-based designs suffer from limited spatial diversity and severe sensitivity to Doppler and multipath effects. While orthogonal time-frequency space (OTFS) modulation offers robustness under high mobility, the impact of spatial node deployment in multistatic OTFS-ISAC has remained largely unexplored. This paper presents the first geometry-aware multistatic OTFS-ISAC framework, in which a triangulation-based cooperative sensing approach is developed for joint target localization and velocity estimation. Closed-form expressions for the localization error covariance are derived under general receiver topologies, revealing that maximizing the triangulation area is fundamental to minimizing estimation error. This leads to a near-optimal deployment strategy based on orthogonal receiver placement and its equivalence to multi-antenna architectures with cubic-order error reduction. To enable reliable tracking of moving targets, a correlated random walk (CRW)-based Kalman filter (KF) framework is integrated into multistatic OTFS-ISAC for active sensing and ISAC. Numerical results demonstrate significant reductions in localization root-mean-square error (RMSE) and communication bit error rate (BER), highlighting the effectiveness of geometry-aware, KF-assisted multistatic OTFS-ISAC in dynamic vehicular environments.

翻译：集成感知与通信（ISAC）是下一代车载网络的一种前景广阔的范式，然而现有的基于正交频分复用（OFDM）的设计存在空间分集有限以及对多普勒和多径效应极为敏感的问题。虽然正交时频空间（OTFS）调制在高移动性下具有鲁棒性，但空间节点部署在多基地OTFS-ISAC中的影响在很大程度上仍未得到探索。本文提出了首个几何感知的多基地OTFS-ISAC框架，其中开发了一种基于三角测量的协同感知方法，用于联合目标定位与速度估计。在一般接收机拓扑下推导了定位误差协方差的闭式表达式，揭示了最大化三角测量面积是减小估计误差的基础。这引出了一个基于正交接收机放置的近似最优部署策略，及其与具有立方阶误差缩减的多天线架构的等价性。为了实现运动目标的可靠跟踪，一个基于相关随机游走（CRW）的卡尔曼滤波器（KF）框架被集成到多基地OTFS-ISAC中，用于主动感知与ISAC。数值结果表明，定位均方根误差（RMSE）和通信误码率（BER）显著降低，突显了几何感知、KF辅助的多基地OTFS-ISAC在动态车载环境中的有效性。