Integrated Communication and Navigation (ICAN) has emerged as a crucial technology for embedding navigation functions into 6G satellite internet waveforms. However, the effective ICAN poses challenges for ICAN-enabled Low Earth orbit (LEO) satellites. Existing methods struggle to balance the requirements of communication and navigation within a single waveform and fail to adequately address the high-dynamic scenarios of LEO satellites. To overcome these challenges, this paper proposes an Orthogonal Frequency Division Multiplexing (OFDM)-based ICAN signal scheme. Leveraging the symmetry of the pilot cross-ambiguity function, it enables precise joint time-frequency channel estimation. Firstly, by directly utilizing the intermediate results of communication, it achieves the delay and Doppler estimation limits while ensuring communication performance. Secondly, an analytical solution for the delay and Doppler estimation error is derived, allowing the direct determination of the minimum pilot allocation under any given accuracy requirements, effectively balancing the trade-off between communication and navigation. Thirdly, field results from an ICAN-enabled LEO satellite (altitude 1100 km) show that the delay and Doppler estimation accuracy can be controlled within 10 cm and 12 Hz under a pilot ratio of 0.089, respectively, with an approximate 58.34% accuracy and 10-15 dB Carrier-to-Noise Ratio improvement over that of traditional Global Navigation Satellite System. These findings offer valuable insights and practical references for the design and implementation of ICAN-enabled LEO satellites.

Image fusion seeks to seamlessly integrate foreground objects with background scenes, producing realistic and harmonious fused images. While existing methods often insert objects directly, adaptive and interactive fusion-requiring contextual adaptation and foreground-background interplay-remains a challenging yet critical task. To address this, we first propose a pipeline for generating high-quality fusion data. By combining iterative in-context learning with existing tools, we curate a diverse cross-scene dataset supporting three core tasks: object integration, replacement, and attribute-referenced editing. Leveraging this, we introduce DreamFuse, a unified diffusion-based approach that jointly optimizes these capabilities. DreamFuse exploits the Diffusion Transformer (DiT) architecture, using its attention mechanism to extract and align foreground-background features for coherent fusion. For flexible control, we incorporate a Positional Affine mechanism, enabling precise spatial and scale adjustments while supporting diverse text-driven fusion. Furthermore, we employ Localized Direct Preference Optimization (L-DPO), refining the model via human feedback to enhance harmony and consistency. Extensive experimental results demonstrate DreamFuse's superiority over state-of-the-art approaches across multiple metrics.

Image fusion seeks to seamlessly integrate foreground objects with background scenes, producing realistic and harmonious fused images. Unlike existing methods that directly insert objects into the background, adaptive and interactive fusion remains a challenging yet appealing task. It requires the foreground to adjust or interact with the background context, enabling more coherent integration. To address this, we propose an iterative human-in-the-loop data generation pipeline, which leverages limited initial data with diverse textual prompts to generate fusion datasets across various scenarios and interactions, including placement, holding, wearing, and style transfer. Building on this, we introduce DreamFuse, a novel approach based on the Diffusion Transformer (DiT) model, to generate consistent and harmonious fused images with both foreground and background information. DreamFuse employs a Positional Affine mechanism to inject the size and position of the foreground into the background, enabling effective foreground-background interaction through shared attention. Furthermore, we apply Localized Direct Preference Optimization guided by human feedback to refine DreamFuse, enhancing background consistency and foreground harmony. DreamFuse achieves harmonious fusion while generalizing to text-driven attribute editing of the fused results. Experimental results demonstrate that our method outperforms state-of-the-art approaches across multiple metrics.

With the rapid development of low-orbit (LEO) satellites, the design of NavCom signals has increasingly become a focal point of research. Among this the orthogonal time-frequency space (OTFS) modulation operates in the Delay-Doppler (DD) domain, effectively mitigating the challenges of highly dynamic channels. Furthermore, the channel estimates obtained can be directly utilized for positioning. However, traditional OTFS is limited to integer delay estimation, which can introduce errors of nearly 100 meters. Most existing approaches have concen-trated on refining the solution algorithms, with accuracy levels approaching their theoretical limits. In this paper, an innovative method is proposed, which superimposes an artificial perturbation on the transmitter signal, causing the values at the receiver to follow a specific distribution and enabling fractional channel estimation after simple processing. This approach expands the theoretical limits of conventional channel estimation by increasing Fisher Information. Moreover, it can be integrated with other algorithms to achieve exceptional performance as an independent enhancement.

This paper proposes a simultaneous transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) segmented symbiotic backscatter non-orthogonal multiple access (NOMA) system, where the STAR-RIS is composed of an enhancing primary signal (EP) zone and a backscatter device (BD) zone. To evaluate the overall system transmission reliability, we derive a tight lower bound of the coexistence outage probability (COP), where an imperfect/realistic successive interference cancellation (SIC) is considered. It is shown analytically that the error floor of the COP in both the near field and far field coverages would appear as long as the residual interference due to SIC is non-negligible, resulting in a diversity order of zero. Such an error floor is mainly dominated by the residual interference parameters, the decoding thresholds, and the power allocation ratios. More importantly, unlike the Gamma approximation approach, the adopted Laplace approach can capture the true diversity order with perfect SIC in the near-field/far-field coverage, which is dominated by the bottleneck number of the STAR-RIS elements belonging to the EP and BD zones, regardless of the dual-hop channel statistics. In addition, it is shown that the COP performance improves with either the quantification order or the concentration parameter of the imperfect channel state information.