ZT-DAC: ML-Driven Dynamic Access Control Policy for Zero Trust in DER Networks
Organizations: Department of Electrical and Computer Engineering, Iowa State University, Ames, IA 50011
Abstract
The increasing penetration of Distributed Energy Resources (DERs) in the smart grid has significantly expanded the cyber-attack surface. Traditionally, DER networks rely on static role-based access control (RBAC) mechanisms that assume persistent trust after authentication. This limitation enables adversaries to maintain unauthorized privileges and move laterally across DER systems, disrupting grid operations. Although attribute-based access control (ABAC) enables fine-grained and context-aware access decisions, it does not fully eliminate implicit trust. Zero Trust (ZT) principles can be incorporated to enhance ABAC policies. This paper proposes a dynamic access control framework for implementing ZT in DER networks (ZT-DAC). The proposed solution employs a two-stage verification process and integrates an Open Policy Agent (OPA)-based RBAC with a machine learning (ML)-based ABAC to enforce continuous and adaptive trust score assessment. To support realistic DER operations, a DER-specific access control model is developed and used to generate a domain-informed dataset of access requests incorporating both legitimate and adversarial behaviors. Experimental results on the dataset show that the OPA-based decision layer achieves accuracy, with a False Positive Rate of , while the ML model achieves a Mean Absolute Error of and an of in trust-score prediction. The ABAC layer achieves accuracy, with a FPR of , validating its practical feasibility within realistic DER environments.
Figures & tables
| Method | OT/ICS | DER-specific | Cybersecurity | Access Control | Machine Learning | Zero Trust |
| Dynamic RBAC using RL [ 7 ] | ||||||
| ZT Grid-Cloud AC [ 8 ] | ||||||
| BC-Resilient DERMS [ 9 ] | ||||||
| IPFS+BC Sharing [ 10 ] | ||||||
| Adaboost AC for SCADA [ 11 ] | ||||||
| ZT-DAC (Proposed) |
| Role | Ownership | Access Channel | Grid Actions | Cyber Responsibilities |
| DER Owner (Residential) | DER equipment | Local interfaces (Wi-Fi, Zigbee) | None | Delegates control to vendor/aggregator |
| DER Owner (Commercial) | DER equipment | Vendor portals or cellular (4G/5G) | None | DER management |
| DER Vendor (OEM) | None (post-deployment) | Proprietary communication channels | None | Firmware patching (on request) |
| Gateway Vendor | None (post-deployment) | None | None | Gateway patching and updates |
| DER Service Provider | None | APIs and DER control servers (standardized/proprietary) | None | Firmware updates and access control management |
| DER/Gateway Installer | None (post-deployment) | Local interfaces (Wi-Fi, Zigbee) | None | Device configuration and security setup |
| Properties | RBAC | ABAC | ZT-DAC |
| Role-Based Authorization | |||
| Context-Aware Authorization | |||
| Behavior-Aware Trust Evaluation | |||
| Continuous Trust Re-evaluation | |||
| Adaptive Access Control | |||
| Lateral Movement Prevention |
| Scenario | Attributes | Legitimate/Illegitimate |
| Grid operator requests monitor_grid_cyber_events for an authorized DERMS resource. | Legitimate | |
| DER aggregator requests dispatch_der_resources for an assigned DER fleet. | Healthy device, compliant firmware, valid certificate, normal request rate, expected operating hour, and low IP-reputation risk. | Legitimate |
| Residential DER owner requests issue_voltage_setpoints , an operation reserved for the grid operator. | Illegitimate | |
| DER service provider requests rotate_credentials on an authorized gateway using an expired certificate. | Healthy device, compliant firmware, expired certificate, authorized role–action–resource mapping, normal request rate, expected operating hour, and low IP-reputation risk. | Illegitimate |
| Feature | Description |
| client_id | Unique identifier of requesting entity. |
| allow | Ground-truth access decision. |
| role | Operational role of requester. |
| action | Requested operation. |
| resource | Target DER asset or service. |
| health | Device health state. |
| Metrics | Description |
| Mean Absolute Error (MAE) | Avg. difference between predicted and reference trust scores. |
| Root Mean Squared Error (RMSE) | Prediction error with higher penalty for large deviations. |
| Coefficient of Determination ( ) | Variance in trust score explained by the model. |
| Accuracy | Percentage of correct allow/deny decisions. |
| False Positive Rate (FPR) | Legitimate requests incorrectly denied. |
| False Negative Rate (FNR) | Suspicious requests incorrectly allowed. |
| Decision path | Mean | Median | Min | Max |
| OPA path | 5.576 ms | 4.550 ms | 3.150 ms | 20.530 ms |
| ML-only | 3.163 ms | 3.060 ms | 2.363 ms | 10.506 ms |
Access Control
Access control specifies and enforces who or what may access data and perform actions, under which conditions.
See all 77 papers