Maintaining IoT network reliability in mixed-mobility megacities like Ho Chi Minh City (HCMC) presents a formidable challenge due to high stochasticity and complex, non-Euclidean traffic topologies. Traditional monitoring systems predominantly treat sensors as isolated spatial points, frequently failing to differentiate between legitimate traffic congestion and latent cascading sensor anomalies. To address this gap, this paper proposes the Graph-Theoretic Spatiotemporal Transformer (GTST). The core innovation of the GTST framework is the integration of a spectral positional encoding scheme derived from the Laplacian eigenvectors of the transport graph, capturing the network's underlying functional structure. Evaluated using a high-fidelity synthetic dataset comprising 100,000 transactions across 28 urban nodes under extreme stochastic noise, empirical results demonstrate the GTST model prioritizes the mitigation of critical false negatives. It achieves a peak fault recall of 0.96, matching computationally heavy deep-learning baselines with a vastly lower computational footprint. Crucially, the GTST framework accomplishes this with an ultra-low inference time of 0.0061ms per transaction, proving its superior viability and efficiency for real-time edge deployment in noisy urban zones.
Thai Ky Trung Pham, Deshinta Arrova Dewi· International Conference on...· 0 citations
In megacities utilizing mixed-mobility transport networks, Back-Office Systems (BOS) are inundated with heterogeneous sensor data—ranging from NFFFC bus validators to GPS telemetry on ferries. Traditional centralized architectures struggle to reconcile this data when faced with stochastic hardware failures, connectivity loss, or "missed taps," which are prevalent in developing urban environments like Ho Chi Minh City (HCMC). The core scientific contribution of this paper is a fairness-aware decentralized integrity mechanism designed to ensure equitable revenue apportionment. To operationalize this, we employ a custom Multi-Agent Reinforcement Learning (MARL) framework, utilizing a Deep Reinforcement Learning (DRL) module as the primary decision engine. The Java Agent DEvelopment Framework (JADE) and a GraphTheoretic Spatiotemporal Transformer (GTST) are utilized strictly as architectural and methodological enablers. These agents negotiate the validity of incomplete sensor logs, learning to probabilistically impute missing data points rather than discarding them. By integrating a GTST for ground-truth estimation, our approach achieves a "fairness-aware" consensus on revenue allocation. Experimental results on a synthetic dataset of 10,000 mixed-mobility transactions demonstrate that this agent-based approach achieves 100.0% revenue recovery with a manageable processing latency of 560 ms/batch. Furthermore, our decentralized MARL framework yields an imputation accuracy of 91.2% and a spatial Root Mean Square Error (RMSE) of 1.12 km—a 54% reduction compared to standard centralized K-Nearest Neighbors (KNN) baselines— while improving algorithmic fairness from 0.54 (discriminatory) to 1.00 (equitable) across peripheral urban zones.
Thai Ky Trung Pham, Deshinta Arrova Dewi· International Conference on...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
