Motivating Learners in Multiorchestrator Mobile Edge Learning: A Stackelberg Game Approach

Mobile edge learning (MEL) is a paradigm that enables distributed training of machine (ML) models over heterogeneous devices (e.g., IoT devices). Multiorchestrator MEL refers to the coexistence multiple tasks with different datasets, each which being governed by an orchestrator facilitate process. In MEL, performance deteriorates without availability sufficient data or computing resources. Therefore, it crucial motivate become learners and offer their resources, either private receive needed from participate in process task. this work, we propose incentive mechanism, where formulate orchestrators-learners’ interactions as 2-round Stackelberg game participation learners. first round, decide task get engaged in, then second parameters amount for case such utility maximized. We study round analytically derive learners’ optimal strategy. Finally, numerical experiments have been conducted evaluate proposed mechanism.