MANET has been a hot research topic for quite a few years. The principal attractions of MANETs are related to the ease with which they can be deployed due to their infrastructure-less and decentralized nature. Applications for MANET range from military operations and emergency disaster relief, to community networking and interaction between attendees at a meeting. The challenge of MANETs is to design dequate security schemes for these innovative networks. A challenging MANET security problem is the issue of the presence of malicious nodes; and these malicious nodes can disrupt the network traffic, choke the bandwidth, cause network congestion and consume unnecessary resources like the limited CPU processing capacity, memory and battery power of individual nodes by making byzantine type attacks. In this paper, we propose a comprehensive security scheme by which these malicious nodes can be easily detected. Here, we use our statistical mechanism that uses the statistics of the activities of the nodes and use a decentralized approach to detect these kinds of malicious nodes. Here, we also simulate our decentralized statistical analysis based intrusion detection scheme over routing protocol for detecting malicious node(s) and analyze the performance and efficiency of our algorithm over previous routing protocol.