.png){kind=icon}
.png){kind=icon}
Return
Multimodal Mobile Edge Computing: Multi-Objective Optimization With Synchronization Constraint
.png){kind=icon}
PRE
AI
DOI：10.1109/TMC.2025.3646465.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Emerging multimodal systems present new requirements for mobile edge networks to handle multimodal data. In this paper, a novel multimodal mobile edge computing (MEC) framework is proposed, which synchronizes the multimodal data acquisition, communication, and computation to ensure both consistency and efficiency. The key objective is to simultaneously maximize multimodal data throughput and minimize the energy consumption of mobile terminals (MTs) under synchronization and resource constraints. A multi-objective optimization (MOP) is formulated, where the sensor activation time, computation offloading, and resource allocation are jointly optimized. To solve this nonconvex problem, a dual-layer Lagrangian multiplier method (D-LMM) is developed. It decouples the optimization into an upper-level throughput maximization and a lower-level energy minimization. The former is converted into a convex problem via quadratic transformation, yielding a stationary solution for sensor activation times, while the latter is solved by alternating optimization. The D-LMM algorithm is proven to converge to a local optimum. Simulation results verify that the proposed framework significantly improves throughput and reduces MT energy consumption. The synchronization-aware multimodal coordination further ensures sufficient data collection and robust performance across varying network scales and resource conditions, enabling reliable downstream operations.
Keywords:
Mobile edge computing
multimodal data
task offloading
resource allocation
Journal
IF:
9.2
Papers:
5.6K
Citations:
1.8W
View more organizations