Semantic communication technology conserves spectrum resources, while unmanned aerial vehicle (UAV)-mounted reconfigurable intelligent surfaces (RIS), a key component of the low-altitude intelligent networks, enhance communication flexibility. Combining these technologies improves wireless performance. However, due to the openness of wireless channels, high-quality communication links are vulnerable to eavesdropping, which compromises system security. Additionally, maritime communication faces challenges such as dynamic channel variations, and high-altitude UAVs struggle to locate sea surface users and eavesdroppers. To address these issues, we propose a secure UAV-RIS-assisted communication scheme that integrates semantic communication, and integrated sensing and communications (ISAC). This scheme maximizes the secrecy semantic rate, reduces communication and computation energy consumption, and ensures constraints on sensing spectral efficiency and semantic accuracy. We jointly optimize UAV-RIS trajectories, RIS phase shifts, spectrum allocation, and the average number of semantic symbols to enhance security under eavesdropping attacks. We propose an approach that integrates semantic communication, the multi-agent softmax deep double deterministic policy gradient, and the multi-agent dueling deep Q-network (S-MA-SD5), effectively supports UAV-RIS-assisted communication in maritime environments. Performance evaluations reveal that the proposed method exceeds existing methods, achieving higher security semantic rates and lower energy consumption, thereby significantly enhancing security and efficiency in the low-altitude intelligent maritime networks.

With the high flexibility of supporting resource-intensive and time-sensitive applications, unmanned aerial vehicle (UAV)-assisted mobile edge computing (MEC) is proposed as an innovational paradigm to support the mobile users (MUs). As a promising technology, digital twin (DT) is capable of timely mapping the physical entities to virtual models, and reflecting the MEC network state in real-time. In this paper, we first propose an MEC network with multiple movable UAVs and one DT-empowered ground base station to enhance the MEC service for MUs. Considering the limited energy resource of both MUs and UAVs, we formulate an online problem of resource scheduling to minimize the weighted energy consumption of them. To tackle the difficulty of the combinational problem, we formulate it as a Markov decision process (MDP) with multiple types of agents. Since the proposed MDP has huge state space and action space, we propose a deep reinforcement learning approach based on multi-agent proximal policy optimization (MAPPO) with Beta distribution and attention mechanism to pursue the optimal computation offloading policy. Numerical results show that our proposed scheme is able to efficiently reduce the energy consumption and outperforms the benchmarks in performance, convergence speed and utilization of resources.