Lidar point cloud synthesis based on generative models offers a promising solution to augment deep learning pipelines, particularly when real-world data is scarce or lacks diversity. By enabling flexible object manipulation, this synthesis approach can significantly enrich training datasets and enhance discriminative models. However, existing methods focus on unconditional lidar point cloud generation, overlooking their potential for real-world applications. In this paper, we propose SG-LDM, a Semantic-Guided Lidar Diffusion Model that employs latent alignment to enable robust semantic-to-lidar synthesis. By directly operating in the native lidar space and leveraging explicit semantic conditioning, SG-LDM achieves state-of-the-art performance in generating high-fidelity lidar point clouds guided by semantic labels. Moreover, we propose the first diffusion-based lidar translation framework based on SG-LDM, which enables cross-domain translation as a domain adaptation strategy to enhance downstream perception performance. Systematic experiments demonstrate that SG-LDM significantly outperforms existing lidar diffusion models and the proposed lidar translation framework further improves data augmentation performance in the downstream lidar segmentation task.

The dual-frequency Global Positioning System has proven to be an effective means of measuring the Earth's ionosphere and its total electron content (TEC). With the advent of multifrequency signals from more Global Navigation Satellite Systems (GNSSs), the opportunity arises to construct many more ionosphere-sensing combinations of GNSS data. With such diversity, various estimable ionospheric delays with differing interpretations (and of different precision) can be formed. How such estimable ionospheric delays should be interpreted, and the extent to which they contribute to the precision with which the unbiased TEC can be estimated, are the topics of this paper. Based on multifrequency GNSS code-only, phase-only, and phase-and-code data, we derive the closed-form solutions of different types of ionospheric observables that each can serve as input of an externally provided ionospheric model for TEC determination. Within such a general least-squares framework, we generalize the widely used phase-to-code levelling technique to its multifrequency version. We also show that only certain specific linear combinations of the observables contribute to the TEC solutions. As a further improvement of the multifrequency GNSS-derived TEC solution, we propose and study the usage of an array of GNSS antennas. Analytical solutions, supported by numerical examples, of this array-based concept are presented, together with a discussion on its relevance for TEC determination. This concerns the roles of time averaging and time differencing, of integer ambiguity resolution, and of the number of frequencies and number of array antennas in determining TEC.