Scholarlink

<p>With the cyber-physical systems championing modern smart grids, securing real-time energy data and transactions is of the essence. Traditional methods such as Signature-Based Intrusion Detection Systems (SB-IDS) are static, thereby missing zero-day threats and incapable of dynamic adaptation, achieving only ~ 85–88% accuracy with false positive rates above 8–10% and incurring &gt; 20% energy overhead on edge devices. In view of this, we propose a blockchain-enhanced cybersecurity framework, where anomaly detection is performed by Graph Convolutional Networks (GCN) and intelligent threat classification by XGBoost. The framework was validated on a 5-bus microgrid (100&#xa0;kW solar, 75&#xa0;kW wind, 250 kWh storage) co-simulated in MATLAB-Simulink with NS-3 network emulation, using over 100,000 labeled attack samples (spoofing, replay, DoS, data injection). Our proposed system is capable of achieving an intrusion detection accuracy of 96.7%, significantly brings down the false positive rate to 2.1%. It can also increase the blockchain transaction throughput to 421 TPS with a latency of just 1.43s, compared to ≤ 20 TPS and &gt; 5&#xa0;s latency reported for Ethereum and Hyperledger-based baselines. Optimized encryption and lightweight models lead to a reduction of cybersecurity-related energy overhead from 23.1% (SB-IDS baseline) to 18.6% measured at edge nodes, thereby allowing energy-efficient secure transactions. Blockchain smart contracts ensure decentralized energy exchange free from any possible manipulation and ML-driven detection adapts dynamically with the evolution of cyber threats. The resilience of the model has been validated with performance evaluation in MATLAB-Simulink for grid simulation and NS-3 for network-level cyber-attack emulation. The proposed framework opens up more opportunities for adaptivity, energy efficiency, and security when compared to SB-IDS, thereby forming a robust base for next-generation smart grid protection.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Energy Efficient Cybersecurity and Blockchain-Enhanced Data Privacy in Smart Grid Infrastructure for Secure Energy Transactions

  • 期刊论文

  • 发表日期：

    2026年3月26日

查看全文
  • B. Srinivasarao,
  • Shaik Reddi Khasim,
  • M. Lavanya,
  • K. Antony Sudha,
  • Gaurav Vishnu Londhe,
  • Satish SamptaroSalunkhe

摘要

With the cyber-physical systems championing modern smart grids, securing real-time energy data and transactions is of the essence. Traditional methods such as Signature-Based Intrusion Detection Systems (SB-IDS) are static, thereby missing zero-day threats and incapable of dynamic adaptation, achieving only ~ 85–88% accuracy with false positive rates above 8–10% and incurring > 20% energy overhead on edge devices. In view of this, we propose a blockchain-enhanced cybersecurity framework, where anomaly detection is performed by Graph Convolutional Networks (GCN) and intelligent threat classification by XGBoost. The framework was validated on a 5-bus microgrid (100 kW solar, 75 kW wind, 250 kWh storage) co-simulated in MATLAB-Simulink with NS-3 network emulation, using over 100,000 labeled attack samples (spoofing, replay, DoS, data injection). Our proposed system is capable of achieving an intrusion detection accuracy of 96.7%, significantly brings down the false positive rate to 2.1%. It can also increase the blockchain transaction throughput to 421 TPS with a latency of just 1.43s, compared to ≤ 20 TPS and > 5 s latency reported for Ethereum and Hyperledger-based baselines. Optimized encryption and lightweight models lead to a reduction of cybersecurity-related energy overhead from 23.1% (SB-IDS baseline) to 18.6% measured at edge nodes, thereby allowing energy-efficient secure transactions. Blockchain smart contracts ensure decentralized energy exchange free from any possible manipulation and ML-driven detection adapts dynamically with the evolution of cyber threats. The resilience of the model has been validated with performance evaluation in MATLAB-Simulink for grid simulation and NS-3 for network-level cyber-attack emulation. The proposed framework opens up more opportunities for adaptivity, energy efficiency, and security when compared to SB-IDS, thereby forming a robust base for next-generation smart grid protection.