A Novel Decentralized Analytical Methodology for Cyber Physical Networks Attack Detection

In many functional implementations of considerable engineering significance, cyber physical solutions have recently been developed where protection and privacy are essential. This led to the recent increase in interest development advanced emerging technology for anomaly intrusion detection. The paper suggests a new frame distributed blind detection by modelling sensor measurements as graph signal using statistical features intrusion. graphic similarity matrices is generated measured data sensors well proximity completely take account underlying network structure. scope collected modelled on random field Gaussian Markov required precision matrix can be determined adjusting called Laplacian matrix. For research statistics, suggested technique based modified Bayesian probability ratio test closed-form expressions derived. end, time analysis actions calculated computing Bhattacharyya distance at consecutive times among measurement distributions. Experiments carried out, evaluated equate efficiency proposed system modern method. findings indicate value better than that offered other existing systems via frame.