Wireless ad hoc networks with additional features of dynamic spectrum access (DSA) are called cognitive radio networks (CRNs) which have the ability to overcome spectrum shortage problem. It is capable to construct collaborative clusters in a dynamic environment to achieve an interference resilience and minimum communication overhead. In this work, a clustering algorithm based on the golden ratio optimization method (GROM) and fuzzy clustering are proposed. The GROM, formulated by the well-known mathematician Fibonacci, is used to select eligible cluster heads in the network. The objective of the optimization fitness function is to choose cluster head by considering the parameters of a higher number of free channels, node degree, remaining energy, speed, and distance to the fusion center. In clustering, fuzzy logic is applied to increase the number of common channels, and it also reduces the number of overlap channels among clusters which increase the cluster stability. The numerical evaluation shows considerable improvement of clustering performance achieved by the proposed clustering when compared with other related works. © 2020 John Wiley & Sons Ltd.