A self-repairing routing algorithm for flying ad hoc networks based on routing quality prediction

The transient instability of Flying Ad Hoc Networks(FANETs), caused by high node mobility and frequent topology changes, leads to frequent link breakage, which in turn increases end-to-end latency, raises energy consumption, and reduces the packet delivery ratio. Regarding the aforementioned problems, existing studies on link quality evaluation and the repair of broken links are still insufficient. To address these challenges, a self-repairing routing algorithm for FANETs based on routing quality prediction is proposed. The algorithm consists of three integrated components: a dynamic routing quality assessment model that evaluates link quality by fusing six routing indicators through AHP and entropy weighting with nonlinear interaction terms, a link transient stability prediction framework that combines Kalman filter-based state estimation with LSTM-based residual learning to forecast link quality degradation, and a sector-constrained self-repairing mechanism that proactively replaces at-risk links before they break. Compared with existing algorithms, the proposed algorithm achieves a competitive packet delivery ratio along with reduced end-to-end delay and reduced energy consumption, effectively enhancing the transient stability of FANETs.