Emerging mega-constellations with numerous Low Earth Orbit (LEO) satellites actively provide pervasive Internet services worldwide, which are usually considered crucial components of Non-Terrestrial Networks (NTNs). However, the high mobility and limited coverage of LEO satellites introduce frequent handovers, causing network interruptions and degrading Quality of Service (QoS). While many efforts have been made to alleviate the impact of handovers on service provisioning from NTNs, they usually assume channel conditions are pre-determined and remain unchanged as satellites move, which is different from real situations and thus may experience significant performance degradation compared to theoretical analysis. In this paper, we proposeOracleto promise QoS-aware service provisioning in NTNs under dynamic channel conditions. Specifically, we mathematically formulate a channel model to characterize dynamic channel conditions in NTNs and develop a QoS maximization problem considering handover frequency and transmission capacity. To accommodate the dynamic nature of NTNs, we introduce a Model Predictive Control (MPC)-based controller to predict future network status and generate control strategies correspondingly, and leverage Digital Twin (DT) for real-time network status consideration. For higher efficiency, we further employ Generative Artificial Intelligence (GAI) with a safe transfer learning-based framework to enhance model adaptivity to environmental uncertainties and ensure feasible control decisions in real-world NTNs. Extensive simulation results under the real-world constellation demonstrate thatOraclecan enhance up to$3\times $QoS during online service provisioning.

Emerging Low-Earth Orbit (LEO) mega-constellations face challenges such as limited bandwidth and highly variable user demand, which can degrade network performance and lead to inefficient satellite resource utilization. One promising solution is to enable Content Delivery Networks (CDNs) within LEO satellites by deploying cache-equipped satellites. However, many existing approaches rely on inter-satellite links, which are not widely used in practice and are typically activated only when terrestrial ground station coverage is insufficient. Furthermore, the dynamic coverage patterns of satellites and diverse regional content preferences add to the complexity of efficient CDN deployment in space. To address these challenges, we proposeSpaceCache+, a satellite-based CDN framework. We introduce a new metric,user benefit, that jointly captures user coverage and latency reduction to assess the effectiveness of cache satellite deployment. Recognizing that deployment typically occurs incrementally, we formulate theUser Benefit-centric Cache Satellite Deploymentproblem and design an efficient heuristic solution. To enhance content placement, we also propose a cache replacement policy based on zero-shot meta-learning, which adapts to both regional content popularity and satellite mobility. We evaluate the performance ofSpaceCache+using real-world constellation settings with CDN traces. Compared with benchmark strategies,SpaceCache+improves user benefit and cache hit ratio by up to 66.29% and 77.12%, respectively.

In recent years, deep or reinforcement learning approaches have been applied to optimise investment portfolios through learning the spatial and temporal information under the dynamic financial market. Yet in most cases, the existing approaches may produce biased trading signals based on the conventional price data due to a lot of market noises, which possibly fails to balance the investment returns and risks. Accordingly, a multi-agent and self-adaptive portfolio optimisation framework integrated with attention mechanisms and time series, namely the MASAAT, is proposed in this work in which multiple trading agents are created to observe and analyse the price series and directional change data that recognises the significant changes of asset prices at different levels of granularity for enhancing the signal-to-noise ratio of price series. Afterwards, by reconstructing the tokens of financial data in a sequence, the attention-based cross-sectional analysis module and temporal analysis module of each agent can effectively capture the correlations between assets and the dependencies between time points. Besides, a portfolio generator is integrated into the proposed framework to fuse the spatial-temporal information and then summarise the portfolios suggested by all trading agents to produce a newly ensemble portfolio for reducing biased trading actions and balancing the overall returns and risks. The experimental results clearly demonstrate that the MASAAT framework achieves impressive enhancement when compared with many well-known portfolio optimsation approaches on three challenging data sets of DJIA, S&P 500 and CSI 300. More importantly, our proposal has potential strengths in many possible applications for future study.

Abstract Differential evolution (DE) has been proven as a simple yet powerful meta-heuristic algorithm on tackling continuous optimization problems. Nevertheless most existing DE methods still suffer from certain drawbacks including the use of ineffective mechanisms to adjust mutation strategies and their control parameters that may possibly mislead the search directions, and also the lack of intelligent guidance and reset mechanisms to escape from local optima. Therefore, to enhance the adaptability of DE-based search frameworks and the robustness on optimizing complex problems full of local optima, an adaptive and guided differential evolution (AdaGuiDE) algorithm is proposed. Essentially, the adaptability of the AdaGuiDE search framework is enhanced by three schemes to iteratively refine the search behaviour at two different levels. At the macroscopic level, the AdaGuiDE search framework revises the existing adaptive mechanism for selecting appropriate DE search strategies by counting the actual contributions in terms of solution quality. In addition, the adaption strategy is extended to the microscopic level where a penalty-based guided DE search is employed to guide the search escaping from local optima through temporarily penalizing the local optima and their neighborhood. Furthermore, a systematic boundary revision scheme is introduced to dynamically adjust the search boundary for locating any potential regions of interest during the search. For a rigorous evaluation of the proposed search framework, the AdaGuiDE algorithm is compared against other well-known meta-heuristic approaches on three sets of benchmark functions involving different dimensions in which the AdaGuiDE algorithm attained remarkable results especially on the high-dimensional and complex optimization problems. More importantly, the proposed AdaGuiDE framework shed lights on many possible directions to further enhance the adaptability of the underlying DE-based search strategies in tackling many challenging real-world applications.