The energy expenditure of underwater sensor nodes significantly exceeds the expenditure of terrestrial sensor nodes due to the challenging conditions of underwater acoustic communication channels. Maintaining and recharging sensor node equipment in underwater wireless sensor networks is particularly difficult due to the complexity of the underwater environment, thereby rendering energy consumption a pivotal concern. Clustering algorithms are regarded as a promising solution to address this challenge. By employing a K-means++-based clustering algorithm for more stable clustering of sensor nodes and taking into account pertinent factors involving node location, residual energy, and node degree in selecting the cluster’s head node, this study targets minimizing network energy consumption over the cluster establishment stage. During the data transfer stage, we prioritize network security and transmission efficiency by implementing a routing strategy that combines single-hop and double-hop propagation paths based on trust security models to mitigate risks from network attacks and black hole nodes. We also introduce a lightweight learning layer for per-link delivery and CH ranking, plus an unsupervised anomaly gate to down-weight suspicious nodes; both fuse with trust and fairness while leaving the physical and energy models unchanged. Simultaneously, AUVs are utilized to anchor sink nodes and directly transmit collected information from CHs. The results of the simulation in MATLAB indicate that the suggested routing protocol KSERP successfully balances and lessens the total energy consumption of the network while offering protection from malicious node attacks.

Forward-looking sonar (FLS) compresses elevation information in 2-D imaging. This makes the recovery of elevation-related structure from a single FLS image an underdetermined problem. Existing pseudo front view (PFV)-based single-frame methods have shown potential, but most still rely on generic isotropic 2-D encoding so known range-azimuth directional characteristics are not sufficiently integrated into PFV recovery. To address this limitation, this letter proposes a direction-aware structured inversion method for single-frame FLS elevation recovery. The method enhances PFV recovery through direction-decoupled feature extraction, azimuth-preserving encoding, and target-aware directional modulation. This improves the representation of geometry-related structural cues in elevation-compressed acoustic observations. Experiments on two public datasets show that the proposed method outperforms existing PFV-based methods and produces more accurate PFV predictions and 3-D acoustic inversion results. Qualitative results on real deep-sea FLS data further support its practical applicability by showing response-consistent 3-D acoustic inversion in measured sonar scenes.