Coordinated Mitigation of Low-Magnitude False Data Injection Attacks on Distribution System State Estimation (DSSE) Using Distributed Ledger Technology and Attention-BiLSTM Models

• Abiodun Okunola

  Ladoke Akintola University Technology

  Author

Distribution System State Estimation (DSSE) serves as a cornerstone for situational awareness in modern active distribution networks, yet its reliance on heterogeneous measurement data renders it vulnerable to stealthy false data injection attacks (FDIAs) that evade conventional bad data detection mechanisms. This study addresses the critical gap in detecting coordinated low-magnitude FDIAs that individually appear benign but collectively destabilize DSSE accuracy. We propose an integrated framework combining an attention-enhanced bidirectional long short-term memory (BiLSTM) network for spatiotemporal anomaly detection with a permissioned distributed ledger technology (DLT) layer for immutable logging and consensus-based attack attribution. The methodology employs the ICS Power Grid dataset augmented with synthetically coordinated attack scenarios, using a three-layer BiLSTM architecture with multi-head attention and layer normalization. Experimental results demonstrate that the proposed Attention-BiLSTM model achieves 89.4% detection accuracy with 91.2% precision and 87.6% recall, outperforming conventional LSTM, CNN-LSTM, and isolation forest baselines. The DLT integration ensures tamper-proof storage of detection alerts with 99.97% integrity verification success. These findings establish a scalable, resilient architecture for securing distribution system monitoring against coordinated cyber-physical threats, with practical implications for utility operators and regulatory frameworks governing critical infrastructure protection.

• pdf

Copyright (c) 2026 Abiodun Okunola (Author)

This work is licensed under a Creative Commons Attribution 4.0 International License.