Abundant datasets are critical to train models based on deep learning technologies for ship detection applications. Compared with optical images, ship detection based on synthetic aperture radar (SAR) (especially the high-Earth-orbit spaceborne SAR launched recently) lacks enough training samples. A novel cross-domain attention GAN (CDA-GAN) model is proposed for optical-to-SAR translation, which can generate high-quality SAR amplitude training samples of a target by optical image conversion. This high quality includes high geometry structure similarity of the target compared with the corresponding optical image and low background noise around the target. In the proposed model, the cross-domain attention mechanism and cross-domain multi-scale feature fusion are designed to improve the quality of samples for detection based on the generative adversarial network (GAN). Specifically, a cross-domain attention mechanism is designed to simultaneously emphasize discriminative features from optical images and SAR images at the same time. Moreover, a designed cross-domain multi-scale feature fusion module further emphasizes the geometric information and semantic information of the target in a feature graph from the perspective of global features. Finally, a reference loss is introduced in CDA-GAN to completely retain the extra features generated by the cross-domain attention mechanism and cross-domain multi-scale feature fusion module. Experimental results demonstrate that the training samples generated by the proposed CDA-GAN can obtain higher ship detection accuracy using real SAR data than the other state-of-the-art methods. The proposed method is generally available for different orbit SARs and can be extended to the high-Earth-orbit spaceborne SAR case.

The cooperative navigation of unmanned aerial vehicle (UAV) cluster has become a popular topic in military and civilian applications. Ultra-wideband (UWB) technology provides a precise and stable practical solution for distance-based cooperative positioning. To develop and evaluate the cooperative navigation system for UAVs, this paper utilizes eight fixed base stations and seven mobile UAVs to collect measurement results under four different flight formations in both indoor and outdoor environments. To eliminate fluctuations in UAV control, predefined flight trajectories were used, and multiple measurements were collected at each trajectory point for technical validation. The resulting data supports the development and validation of distance-based multi-target navigation algorithms, the evaluation of formation maintenance and collision avoidance algorithms for multiple UAVs, and also serves as a source for navigation evaluation and the development of multi-source fusion cluster positioning algorithms.

Current cooperative positioning schemes are complex, and their performances vary, urgently necessitating evaluation. In view of the specific characteristics of Cooperative Positioning Systems (CPS) in terms of physical data, decision-making factors, and the professionalism of decisions, this study aims to construct a reasonable Multi-Criteria Cooperative Positioning Decision-Making Evaluation (MCCPDM) framework. The study applies Generative Adversarial Networks (GAN) to the decision-making process, proposes a weighted fusion of Pythagorean fuzzy sets and hesitant fuzzy sets, and provides a method for determining the weights of decision-making factors that combines objective physical entropy information with subjective multi-level judgments. An MCCPDE method integrating Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) is developed under the weighted Pythagorean hesitant fuzzy environment. Finally, through experiments evaluating cooperative positioning schemes, the proposed framework is proven to be highly effective and practical in cooperative positioning evaluation issues, and the method presented in this paper surpasses the best baseline method by 2 orders of magnitude in terms of optimal ranking proximity.