Research on Information Security Protection in Power Scheduling Networks Based on Distributed Intrusion Tolerance Technology

With the increasing digitization of power systems, ensuring the security and reliability of power dispatch networks has become critically important. This study proposes an innovative information security protection algorithm based on distributed intrusion tolerance technology. The algorithm integrates dynamic data collection and machine learning methods to predict potential faults and optimize resource allocation, thereby enhancing the system's fault tolerance. Experimental results demonstrate that the virtual machine layer exhibits the shortest and most stable fault detection time, followed by the hardware and host layers. Detection accuracy is highest and most stable at the virtual machine level, while the host layer shows lower and more variable accuracy. Load stress tests indicate that the algorithm maintains stable performance under CPU and memory stress, with some fluctuations under mixed I/O stress due to disk I/O failures. These findings provide valuable insights for designing robust fault tolerance mechanisms in power scheduling networks.