Hybrid Metaheuristic-Optimized Multilayer Incentive Mechanism for Time-Dependent Decentralized Mobile Crowdsensing (HMOM-MCS)

Mobile Crowdsensing (MCS) faces persistent challenges in user participation, data freshness (Age of Information), privacy, and scalability in decentralized environments. We introduce HMOM-MCS , a focused framework deliberately centered on three interdependent pillars: (1) the novel Hybrid Bald Eagle Search–Gazelle Optimizer (BES-GO) that efficiently solves the non-convex, time-dependent, multi-objective incentive problem with sublinear regret \(O(√Tlog T) \) under standard online-learning assumptions; (2) a strict mobile-edge-fog-cloud architecture that reduces decision-making latency by 30–40% while enabling real-time incentives; and (3) a participation-centric incentive suite (entry lotteries for newcomers, social referral bonuses, and reputation multipliers) that works synergistically with federated-learning-based user profiling and blockchain settlement. In simulation, and under the behavioral assumptions described in the paper, this design achieves 21–52% higher social welfare, raises the participation rate from below 60% to 85%, attains 98% task completion, lowers information latency by 16–25%, and reduces energy consumption by 20–25% compared to state-of-the-art baselines. On the theory side, we establish incentive compatibility for the VCG core augmented with bid-independent bonuses (with the scope of the remaining incentive terms delimited explicitly), individual rationality, budget feasibility, and sublinear regret under standard online-learning assumptions. Extensive simulations on real-world traces (Rome Taxi, Geolife, Foursquare, CRAWDAD) and ablation studies indicate that, under the simulation settings considered, removing any of the three pillars causes statistically significant performance degradation ( p < 0.001). These results are simulation-based: they characterize behavior under our modeling assumptions rather than measured real-world participation, and human-subject validation in a live testbed remains an essential direction for future work. All code and data are publicly accessible at https://github.com/Ali-Hosseini5422/HMOM-MCS-Simulator.