TrustPulse-IoT: Context-Aware Adaptive Access Control for Heterogeneous IoT Networks

Abstract

IoT deployments increasingly combine constrained sensors, gateways, legacy controllers, cloud services, and vendor-managed interfaces. This heterogeneity makes a single perimeter or static access rule insufficient because a device that was legitimate at enrollment may later exhibit an unsafe communication pattern. This paper proposes TrustPulse-IoT, an adaptive security architecture that converts device identity, behavioral stability, protocol compliance, asset criticality, and telemetry freshness into a continuously refreshed trust value at the edge. The trust value is used to select one of four actions—permit, permit-with-restrictions, quarantine, or deny—while a lightweight behavior profile is maintained locally to reduce dependence on centralized raw telemetry. The design is motivated by security gaps identified in the supplied IoT literature, particularly weaknesses involving device identity, access control, segmentation, insecure interfaces, data protection, and limited attention to continuous security state. Unlike a conventional checklist of controls, the proposed model links observation to policy enforcement and feedback. A formal trust function, decision algorithm, threat model, prototype plan, and evaluation protocol are presented. No experimental performance values are claimed; the paper defines a reproducible path for implementation and measurement. The proposed architecture is intended as an incremental research contribution that can be evaluated against a baseline IoT network using detection quality, unauthorized-access prevention, response latency, overhead, and false-positive rate.