While cooperative perception can overcome the limitations of single-vehicle systems, the practical implementation of vehicle-to-vehicle and vehicle-to-infrastructure systems is often impeded by significant economic barriers. Aerial-ground cooperation (AGC), which pairs ground vehicles with drones, presents a more economically viable and rapidly deployable alternative. However, this emerging field has been held back by a critical lack of high-quality public datasets and benchmarks. To bridge this gap, we present Griffin, a comprehensive AGC 3D perception dataset, featuring over 250 dynamic scenes (37k+ frames). It incorporates varied drone altitudes (20-60m), diverse weather conditions, realistic drone dynamics via CARLA-AirSim co-simulation, and critical occlusion-aware 3D annotations. Accompanying the dataset is a unified benchmarking framework for cooperative detection and tracking, with protocols to evaluate communication efficiency, altitude adaptability, and robustness to communication latency, data loss and localization noise. By experiments through different cooperative paradigms, we demonstrate the effectiveness and limitations of current methods and provide crucial insights for future research.

We address the problem of inferring the network link loss rates using end-to-end measurements, which can also be formulated as network tomography. As we have known that most tomography problems are rank-deficit. One kind of method uses multiple probe measurements to acquire more information about the system that may generate much additional overhead; the other method imposes unrealistic assumption on the system. To address the issue that most network tomography methods cannot take into account both accuracy and efficiency, a novel link loss rate inference algorithm is proposed. In this paper, we get all identifiable links and then we utilize the information of these determined links to acquire the global distribution of the system. Moreover we partition all links in the network into several subsets. For each group, nonlinear programming is used to get the optimization solution of link loss rate. Finally, we evaluate our method and two former representative methods by the simulation. The results demonstrate that our method not only reduces the probe costs and the running time to a low level, but also makes a great improvement on the accuracy. Furthermore, our method can also perform well in more congested and large networks.