This study investigates Embodied Virtual Agents (EVAs) driven by Large Language Models (LLMs) as mediators in triadic collaboration where two users with conflicting goals must work towards a shared objective. We developed and evaluated an XR system where pairs (n=24) collaborated on an office design task, assessing the impact of agent presence and embodiment. Our findings reveal that agent embodiment significantly enhanced co-presence, which in turn fostered interactions that led to better perceived collaboration and higher user preference, whereas frequent interactions with the disembodied agent was associated with lower user preference. Conversely, agent presence did not improve task efficiency or satisfaction. Notably, our qualitative analysis revealed that the LLM-driven agent spontaneously adopted emergent facilitative and evaluative mediation strategies that align with collaborating and compromising conflict-resolving modes, showcasing its potential as an adaptive collaborative aid without explicit programming. These results highlight that the value of EVAs in complex collaboration lies in their ability to shape social dynamics and provide nuanced, context-aware mediation, a different form of value than traditional productivity enhancement.

Low Earth Orbit (LEO) satellite networks, characterized by their low latency and extensive coverage, play a pivotal role in the development of future 6 G communication systems. However, due to the dynamic nature of network topology and the challenges associated with real-time perception of link states, the implementation of deterministic routing in LEO satellite networks presents significant difficulties. To address these difficulties, we propose a network calculus-based deterministic routing (NCDR) algorithm. Specifically, we develop a deterministic resource characterization model and design a traffic pre-transmission mechanism that utilizes network calculus theory to calculate the traffic backlog. Additionally, we propose an interruption feedback mechanism to deal with link interruptions. Finally, the NCDR algorithm makes routing decisions aimed at minimizing end-to-end transmission delay and balancing network load. Simulation results demonstrate that the NCDR algorithm significantly outperforms existing algorithms in terms of delay, throughput, and packet loss rate.

With the advantages of low delay, wide coverage, and high throughput, low-Earth-orbit (LEO) satellite networks hold significant potential for establishing globally interconnected networks. However, the large spatial scale of satellite networks leads to lagging link state perception. Moreover, the perception overhead significantly increases with the growing number of satellites. These characteristics have brought great challenges to the routing design of large-scale LEO satellite networks. To address the above challenges, we propose a multi-region joint routing (MRJR) algorithm based on network calculus (NC) theory to achieve low-delay transmission without relying on traditional state perception. Firstly, a multi-region NC model is introduced to efficiently manage satellite networks and decouple the traffic transmission process. Then, we design the MRJR algorithm, which accurately derives link traffic backlogs to acquire real-time link congestion states, thereby calculating the lowest delay routing path. Additionally, an NC timeslot correction mechanism is proposed to ensure the accuracy of traffic backlog calculations. The simulation results demonstrate that the MRJR algorithm outperforms existing routing algorithms in terms of average delay, throughput, and packet loss.

This paper proposes a probabilistic amplitude shaping (PAS) scheme with systematic cascaded staircase and Hamming codes (CSHC). The integration of low-complexity concatenated codes within PAS and the use of the SIHO decoder are enabled by applying the inter-bit independence assumption. This ensures compatibility with high throughput transmission. In addition, the use of a reduced test pattern set in the SIHO decoder reduces the test pattern cardinality by 34.37%. The proposed scheme is experimentally demonstrated by self-homodyne coherent transmission over a 22.5 km 7-core fiber, with 21 wavelength division multiplexing (WDM) channels in the C-band is demonstrated. The total rate of the WDM-space division multiplexing (SDM) system is 21.77 Tbit/s. The experimental results show that the proposed PAS scheme with CSHC provides a gain of more than 8 dB in the back-to-back (BtB) scenario at the pre-FEC BER of 2.88E-2. Moreover, an average gain of 7.26 dB is achieved at the same threshold in WDM-SDM transmission.

Virtual reality (VR) is increasingly adopted for collaborative teaching and learning, enabling immersive and interactive experiences. As Artificial Intelligence (AI) tutors begin to take on roles alongside human instructors, it becomes crucial to understand how their integration influences interaction dynamics and role perception in these settings. This study investigates the role of Embodied Virtual Agents (EVAs) substituting human instructors in virtual training, specifically addressing the previously underexplored issue of EVA appearance consistency with instructors in Human-Agent Teaming (HAT). We recruited 21 participants to compare three conditions: No Agent, Shared-Appearance (SA) EVA, and Unique-Appearance (UA) EVA, where an EVA substitutes for the instructor during temporary absences. We evaluated collaboration efficiency, user perception/preference, and HAT dynamics. Our findings confirm that EVAs significantly enhance task efficiency compared to no support and reveal a key trade-off regarding appearance: SA fosters perceived continuity and trust but risks ambiguity and uncanny effects, while UA provides transparency and role clarity but may disrupt experiential coherence. These results have implications for designing dynamic HAT systems where control may shift. We discuss the benefits and limitations of each approach and offer design recommendations for future mixed-agency interfaces.

We demonstrate a probabilistic shaping (PS) four-dimensional (4D) modulation in self-homodyne coherent transmission system. The 4D modulation is based on inter-symbol amplitude translation (AT) to perform set partitioning. The distribution of constellation points after AT is optimized by de-DC. The parity bits produced by the AT are transmitted with the pilot tone by remapping. In addition, a soft decision for this 4D-PS signal is proposed. An experiment of self-homodyne coherent ultra-high order 4D signal transmission based on two cores of a 7-core fiber is demonstrated with a spectral efficiency of 16.37 bit/s/Hz. The 4D signals with soft decision can provide up to 0.78 bit/symbol and 1.85 bit/symbol gain compared to normal polarization division multiplexing signals and hard-decision 4D signals.

As an important complement to the terrestrial network and an essential component of the future 6G, Low Earth Orbit (LEO) satellite network is expected to provide higher-quality communication services in combination with terrestrial network and attracts widespread research interest. Since the unbalanced distribution of terrestrial services may lead to inter-satellite link (ISL) congestion, balancing the network load has become one of the key issues for LEO satellite networks. In order to prevent ISL congestion in LEO networks, we propose a multi-region division routing algorithm based on fuzzy-shortest-path-first (FSPF-MDR) for LEO satellite networks. We divide the satellite network into multiple small regions and share link information within the regions to achieve congestion avoidance. Simulation results show that, using the proposed algorithm, the computing complexity is reduced significantly with a slightly cost of increasing average hop count. Furthermore, it is able to achieve congestion prevention while using fewer resources. With different scales of LEO networks, the computing complexity can be reduced by 45% to 70%.

Multi-output regression estimation aims at mining a vector-valued function from multi-dimensional input vector to multi-dimensional output vector. However, the output variables may be correlative. It is desirable to develop a multi-dimensional regression model taking advantage of the possible correlations. Therefore, this paper proposes a novel multi-output support vector regression model via double regularization. For each output variable, we first introduce an influential level vector with the dimensionality equal to the one of the input vector, in order to characterize the correlation between this variable and other output variables. As a second regularization term, 2-norms of all influential level vectors are then added into the objective function. Each influential level vector is also considered in constraints of our model. Finally, experimental comparisons demonstrate that our proposed model in this paper has a better generalization performance as well as a better robustness.