Abstract-Hello Flood attack is a type of attack in wireless sensor networks. In this attack, the malicious node is able to disturb the security of network by sending periodic hello packets with high signal strength. In this research, an intrusion detection system based on neighborhood is proposed. It is based on a principle that those sensor nodes which are close to each other spatially are tended to have similar behavior. If a node shows a significant difference in its behavior in comparison with other nodes in neighborhood, it is considered as a malicious node. For optimizing, the adaptive method filtering based on Alpha-Beta is used. For our simulation, we have used TOSSIMthe sensor network simulator in TinyOS systems. The simulating result shows the detection method based on neighborhood has a high detection accuracy and low false positive rate. The proposed method also detects collusion attack of malicious nodes.Index Terms-Wireless sensor network, hello flood attack, adaptive intrusion detection, collusion attack.