Semantic Communication-Based Aerial-Maritime Energy Trade-Off for Maritime Mobile Edge Computing Networks Under Jamming

Unmanned aerial vehicles (UAVs)-enabled maritime mobile edge computing (MEC) effectively meets the increasing computational demands in maritime environments. However, the limited energy, computing resources, and bandwidth of wireless communication networks present significant challenges for timely computation and offloading efficiency. Moreover, the extensive line-of-sight (LoS) in maritime communication links makes MEC networks more vulnerable to malicious jamming than terrestrial networks. To this end, we investigate a semantic-aware aerial-maritime MEC network impacted by a malicious jammer, in which a UAV provides semantic communication (SemCom)-based task offloading services for maritime devices (MDs) with binary offloading. To reveal the energy trade-off between the MDs and UAV, our objective is to minimize their total weighted energy consumption by optimizing computing resources, UAV trajectory, and task offloading indicator, subject to task completion time constraints and semantic similarity constraints, in the presence of jamming. The formulated problem is a mixed binary variable strongly nonconvex problem, making it difficult to find the optimal solution. To tackle this problem, we propose a double-loop penalty-based alternative energy minimization (P-AEM) algorithm that achieves a high-quality suboptimal solution, therein exact penalty method (EPM) is employed to address binary constraints. To address the variables coupling and nonconvexity of the subproblems, we design a hybrid hierarchical difference of two convex (HD.C.) and logarithmic transformation optimization framework, followed by the successive convex approximation (SCA) technique to achieve a suboptimal solution. Simulation results demonstrate that the proposed scheme significantly reduces the total energy consumption compared with benchmark schemes and reveals the aerial-maritime energy trade-off.