There are inherent vulnerabilities that are not easily preventable in the mobile Ad-Hoc networks.To build a highly secure wireless Ad-Hoc network,intrusion detection and response techniques need to be deployed;The intrusion detection and cluster-based Ad-Hoc networks has been introduced,then,an architecture for better intrusion detection based on cluster using Data Mining in wireless Ad-Hoc networks has been shown. A statistical anomaly detection approach has been used.The anomaly detection and trace analysis have been done locally in each node and possibly through cooperation with clusterhead detection in the network.