An Adhoc network is a mix of mobile nodes that functions without centralized infrastructure. Every mobile node works as host and routers. In addition, it also transmit packets to additional mobile nodes in the network that are not within the direct broadcasting range. Mobile ad-hoc networks are readily subject to different network layer assaults such as black hole, wormhole, as well as DOS attack. Wormhole attack is one of the serious assaults in MANET. Wormhole attacker gets the packets at any given site in the network and interrupts the flow of packet through tunneling them to a different location. In this research, enhanced method is described against these wormhole assaults in a MANET. Particularly, due to their incapacity to maintain node dependability, MANETs are susceptible to routing assaults like wormhole attacks. Since wormhole attacks generally do not immediately destroy networks, discovering them may be tricky. In response, we offer a unique multiple verification-based wormhole identification approach that harnesses the peculiarities of such assaults. The suggested technique assesses the credit of all nodes based on a trust mechanism. The trust levels for suspicious nodes are decreased throughout routing; those having trust levels below a specific threshold are deemed malevolent. This trust system was created utilizing AODV inspired Deep reinforcement learning, which enhances the accuracy of the algorithm over time. Simulation studies in which the suggested technique was applied to current routing algorithms in a highly populated environment were undertaken; the rate of traffic going via pathways containing malevolent nodes was dramatically decreased. Existing solutions employ Quality of Service (QoS) for whole network to identify attacks. Our technique takes utilize of the packet delivery rate and round trip duration for each node, and it also identifies active and passive assaults. Thus, the entire detection of a wormhole attack is attainable using the provided approach