Although it is a fundamental technique for autonomous navigation in GNSS-denied environments, tightly-coupled multi-sensor fusion has advanced faster than conventional narrative assessments. Using a PRISMA-guided review, bibliometrics, and Latent Dirichlet Allocation (LDA) topic modeling ($K=8, C_v=0.62$), we analyze a corpus of \Nincluded{} included publications (2015–2026) to give a data-driven synthesis of the domain. Resilience appears as an increasingly explicit topic across the corpus: more recent publications highlight degeneracy, outliers, and event-driven sensing inside state-estimation pipelines, in contrast to previous work that concentrated on nominal robustness. The literature from 2024 to 2025 particularly shows four trends (2026 is a half year in our export).:(i) measurement-domain integration of 5G, WiFi Round-Trip Time (WiFi-RTT), and ultra-wideband (UWB) signals; (ii) event-based and neuromorphic vision for high-dynamic-range scenarios; (iii) decentralized swarm cooperative localization; and (iv) semantic-cognitive map integration for managing geometric degeneracy. We also find a reporting gap: a lot of studies don't include computational performance data, which makes real-world validation more difficult. The analysis scripts and machine-readable corpus metadata are available upon request. Lastly, using three synthesis tools—a conceptual degradation--response framework for navigation resilience, an empirically supported taxonomy of degeneracy-mitigation mechanisms, and an engineering decision matrix for choosing sensor--estimator topologies under computational and environmental constraints—we convert the review findings into useful recommendations.
Swarm unmanned surface vehicles (USVs) have become significantly prominent for maritime defense and security operations. Cooperative encirclement task, as part of maritime operations, is a critical capability of autonomous swarm USVs that requires motion stability, robust coordination, and energy efficiency under dynamic conditions. The key challenges in encirclement tasks are that most approaches focus solely on translational and positional motion while neglecting the integration of rotational motion around the target. In addition, convergence speed remains a persistent issue in reinforcement learning-based methods. To address these challenges, this work proposes the integration of Tesla-valve optimization into multi-agent reinforcement learning (TVO-MARL) to improve the encirclement task performance. The novel TVO approach is applied into two baselines of MARL including multi-agent proximal policy optimization (MAPPO), and MAPPO-LSTM (long short-term memory). The proposed methods are assessed in the 3D simulation platform using multiple performance metrics, including cumulative reward, angular coverage, rotation number, coverage variance, trajectory smoothness, and energy consumption. Experimental results show that proposed methods outperform the baseline for the convergence speed in training stage. Moreover, the top-performing trained models demonstrate that MAPPO-TVO significantly improves rotational stability, while MAPPO-LSTM-TVO achieves the highest cumulative reward and angular coverage, along with smoother trajectories and substantially reduced energy consumption. Statistical analysis confirms that these improvements are significant compared to baseline MAPPO and MAPPO-LSTM methods. The results reveal clear trade-offs across encirclement objectives and highlight the complementary benefits of nature-inspired optimization via TVO and temporal policy representation, providing a robust solution for swarm USV encirclement task.