Abstract
In network security, various protocols exist, but these cannot be said to be secure. Many hackers and illegal agents try to take advantage of the vulnerabilities through various incremental penetrations that can compromise critical systems. The conventional tools available for this purpose are not enough to handle things as desired. Risks are always present with dynamically evolving networks and are very likely to lead to serious incidents. This research work has proposed a model to visualise and predict cyber-attacks in complex, multilayered networks. All the available network security conditions and the possible places where an attacker can exploit the system are summarised. The vulnerability-based multi-graph technique for the attacker is presented. Also, an attack graph algorithm is proposed, leading to identifying all the vulnerable paths that can be used to harden the network by placing sensors at the desired locations and is used for vulnerability assessment of Multi-Stage Cyber-Attacks.