This paper investigates time synchronization in distributed sensor networks using the primary synchronization signal (PSS) in Long-Term Evolution (LTE). Two LTE-based time synchronization methods with receiver-to-receiver characteristics have been evaluated in simulations, the passive Scalable Wireless Network Synchronization (SWINS) and the active Reference Broadcast Synchronization (RBS). In addition, small scale hardware experiments were conducted for SWINS. Time synchronization is crucial for many applications, such as power grid monitoring, communication systems, and sensor data fusion. Global Navigation Satellite Systems (GNSS) are currently the state of the art for time synchronization in distributed wireless sensor networks. However, GNSS is vulnerable to jamming and spoofing, which requires alternative methods, e.g., using signals of opportunity. Both evaluated methods achieve accuracy comparable to GNSS in Matlab simulations with high SNR. SWINS performs better in synchronized LTE networks while RBS is superior in unsynchronized networks, which means that the LTE base station transmissions are not synchronous. The disturbance and sensitivity analysis indicates that joint clock offset and position estimation is preferable to sole clock offset estimation when the receiver position uncertainty exceeds 5 and 7 meters for SWINS and RBS respectively. The hardware experiments, using real experimental data, verify the simulation results by showing promising results and potential for real-world application.

This paper considers the problem of detecting and tracking objects in a sequence of images. The problem is formulated in a filtering framework, using the output of objectdetection algorithms as measurements. An extension to the filtering formulation is proposed that incorporates class information from the previous frame to robustify the classification. Further, the properties of the object-detection algorithm are exploited to quantify the uncertainty of the bounding box detection in each frame. The complete filtering method is evaluated on camera trap images of the four large Swedish carnivores, bear, lynx, wolf, and wolverine. The experiments show that the class tracking formulation leads to a more robust classification.