知识驱动的无人机网络频谱资源群智优化研究

With the development of artificial intelligence technology, the intelligent coordinated UAV swarm turns to be an important force in the military and civilian fields such as future wars and emergency rescue. The optimization of communication, especially the spectrum resources, is one of the foundations for realizing the group intelligence of drones. However, existing optimization methods cannot be well adapted to the characteristics of large-scale, strong heterogeneous, high dynamics of the UAV cluster. Therefore, this project aims at the new requirements of the intelligent cooperative UAV group for spectrum resources, and the knowledge-driven group intelligence optimization as the guiding ideology. Based on the group consensus excitation and gather, the deployment of UAV group formation and spectrum resources under large-scale conditions is optimized. The matching degree between spectrum and requirements under strong heterogeneous conditions is improved based on scenario perception such as business and trajectory. Based on the knowledge migration framework, the efficiency of spectrum access decision in high dynamic scenarios is increased. Finally, the stable allocation of resources, collaborative sharing and fast access are realized. In order to achieve breakthroughs in the theory of group intelligence optimization and key technologies of spectrum resource intelligent management, knowledge-driven spectrum resource group intelligence optimization is realized. Overall, this project aims to provide communication foundation and guarantee for the group intelligence of the drone group, and to make positive contributions to the future intelligentization of our army.

随着人工智能技术的发展，智能协同无人机群已经成为未来战争、应急救灾等领域中的重要力量。其中，通信资源的优化，尤其是频谱资源，是实现网络群体智能的重要基础。然而，现有的优化方法不能很好地适应无人机集群大规模、强异构、高动态的特点。本项目瞄准无人机网络中的频谱资源优化问题，以知识驱动的群智优化为思路，基于群体共识的激发和汇聚，优化大规模无人机群编队资源与频谱资源的部署和配置；基于业务、轨迹等情景知觉，提高强异构条件下的频谱资源和业务需求的匹配程度；基于知识迁移框架，增加高动态场景中频谱接入决策效率，实现资源的协同分配，高效共享和快速接入。以期在频谱资源智能管控关键技术方面取得突破，实现知识驱动的频谱资源群智优化，从而为无人机群的智能协同提供通信基础和保障，为推动我军未来信息化智能化建设做出积极贡献。

随着人工智能技术的发展，智能协同无人机群已经成为未来战争、应急救灾等领域中的重要力量。其中，通信网络资源的优化，尤其是频谱资源，是实现网络群体智能的重要基础。本项目以提高无人集群通信网络资源利用效率为优化目标，深入研究了基于博弈论的频谱资源优化方法，取得的主要研究进展和成果包括：.1. 针对无人集群在任务共识和内容共识方面等方面的共同性需求，基于共识汇聚的频谱分配优化思路，研究了基于联盟形成博弈的功率控制和干扰消除、基于内容相似性的多子任务负载均衡等问题，有效消除了集群内部干扰，提升了频谱资源分配效率；.2. 针对无人集群在业务传输、时延、能量等方面的差异化需求，基于情景知觉的频谱共享优化思路，研究了基于计算负载约束下基于业务知觉的频谱-计算联合优化、面向节能的无人机群组播通信资源优化等问题，提高了频谱资源利用效率；.3. 针对无人集群在电磁信号调制识别、辐射源识别等问题，通过基于深度学习算法的特征提取能力，研究了基于电磁感知网络的轻量级信号调制方式识别、基于电磁信号数据集扩充的辐射源识别，有效提升了电磁信号识别能力。.项目结题已在IEEE Transactions on Wireless Communications、IEEE Systems Journal等高水平期刊和会议上发表论文12篇，申请发明专利4项，完成了预期研究目标和预期研究成果。

内容获取失败，请点击重试