User Scheduling for Integrated Communication and Semantic Localization in D-MIMO

We focus on user scheduling for integrated communication and semantic localization in distributed multiple-input multiple-output systems with limited uplink resources. The scheduling metric is based on semantic misclassification risk, linking localization uncertainty to the geometry of the spatial decision. Communication requirements are enforced through probabilistic queue constraints, while remaining resources are allocated according to the predicted reduction in semantic risk. This shifts localization-aware scheduling from accuracy-based allocation toward task-relevant information acquisition. Numerical results show that semantic risk-based scheduling consistently outperforms scheduling based on traditional metrics.