The availability of the high-capacity network, massive storage, hardware virtualization, utility computing, service-oriented architecture leads to high accessibility of cloud computing. The extensive usage of cloud resources causes oodles of security controversies. Black-hole & Gray-hole attacks are the notable cloud network defenseless attacks while they launched easily but difficult to detect. This research work focuses on proposing an efficient integrated detection method for individual and collusion attacks in cloud computing. In the individual attack detection phase, the forwarding ratio metric is used for differentiating the malicious node and normal nodes. In the collusion attack detection phase, the malicious nodes are manipulated the encounter records for escaping the detection process. To overcome this user, fake encounters are examined along with appearance frequency, and the number of messages exploits abnormal patterns. The simulation results shown in this proposed system detect with better accuracy.