It is necessary for securing wireless sensor networks from jamming, since these networks are easily attacked by jammers. Jamming may be made by either external jammer node or internal node, which may become as an adversary in future. In this paper, two approaches are proposed in the cluster-based sensor network to detect the maliciousness level of nodes to secure sensor networks from jamming attacks. First approach detects maliciousness level of nodes using two modules, namely, certification module and monitoring module. Certification module defends the network from the jammers. Monitoring module discovers the sensor nodes that are jammed by a jammer. Second approach uses fuzzy logic for optimizing the jamming metrics to determine the occurrence of jamming accurately. The proposed system achieves 99.58% detection ratio to determine node's maliciousness level. © 2019 John Wiley & Sons, Ltd.