MANET (Mobile Ad hoc Network) is a collection of self-organizing mobile nodes that operate independently of centralized administration or existing infrastructure. MANETs are particularly vulnerable to a range of security risks as a result of this feature. Furthermore, most MANET routing methods are built on the assumption that the network contains no hostile nodes. As a result, various efforts and studies have gone into developing ad hoc networks, and a safe and reliable routing mechanism. Adaptive cuckoo search optimization and entropy-based signature authentication were proposed in this research for safe routing in MANETs. The network’s nodes are first clustered using energy sigmoid-based fuzzy c-means clustering. The nodes in the cluster are then given an entropy-based signature. After that, the link lifetime and path distance are estimated to determine the node’s future location. Adaptive cuckoo search optimization is then used to find further optimized routes. Finally, entropy-based signature verification is used to ensure secure routing. The experimental results show that the suggested secure routing technique surpasses existing techniques in terms of performance.