Wireless sensor nodes remain vulnerable to malicious attacks in hostile environments. Adversarial attacks capturing a few benevolent nodes and creating multiple duplicate identities make the localization process very tricky. Incorrect position estimations often lead to the wrong decision regarding the monitored event. This paper presents a secure localization scheme mitigating such attacks using fuzzy-based anchor mobility control. All nodes periodically broadcast signals over the network using omnidirectional antennas. The anchor each time forms an individual cluster using some neighboring nodes with an $${\text{Received Signal Strength}}$$ Received Signal Strength ($$\mathrm{RSS}$$ RSS ) exceeding a prescribed threshold. The mobility control algorithm keeps the anchor moving along a path that allows cluster formations at two successive anchor positions containing at least the nodes encountered first in the previous case. The current cluster size and the node location with minimal signal strength are two determining factors for the fuzzy controller to explore a new anchor path. The misbehaving nodes are detected and blocked by observing anomalies in node identities within the subsequent two clusters. The anchor contains an adaptive antenna and remains location-aware with a Global Positioning System (GPS) receiver. The anchor determines the $${\text{Angle of Arrival}}$$ Angle of Arrival ($$\mathrm{AoA}$$ AoA ) information of all the signals intercepted in each cluster using such an antenna. Then, the anchor generates and transmits beacon messages to each legitimate node. Upon receipt of two beacon messages, the nodes localize themselves autonomously and switch to power-saving mode. The simulation results corroborate its competence comparable to existing methods for localization accuracy and intrusion detection precision.