Robustness analysis of CPPS against malware attacks under malware incubation and attack–defense confrontation

The integration of advanced technologies in Cyber-Physical Power System (CPPS) has increased the vulnerability to cyber threats. Consequently, many network science-based studies have sought to analyze the robustness of CPPS against Advanced Persistent Threats (APTs). Compared with previous studies, our framework incorporates the malware incubation period, incomplete attacker information, coupling patterns, and reconnaissance-evasion behavior, thereby covering a broader spectrum of scenarios. Based on our framework, we first quantify the tradeoff between incubation duration and attack success rate to identify the optimal attack launch time. Subsequently, we investigate how system robustness is affected by critical parameters across various scenarios. Simulation experiments reveal that the attackers’ incomplete information delays the optimal launch time. Disassortative coupling patterns can mitigate malware impact and enhance CPPS robustness. Furthermore, results demonstrate two distinct defensive resource allocation strategies for optimal CPPS robustness. In the disassortatively coupled CPPS, defenders must continuously adjust the deployment of defensive assets as the relative attack–defense budget shifts, whereas under assortative coupling a quasi-static deployment of defensive assets suffices to attain optimal robustness. These insights provide cost-effective guidelines for strengthening the robustness of CPPS.