Cognitive radio (CR) is an intelligent radio that can coexist with licensed users without affecting their QOS. It is very likely that 5G devices will have CR functionalities. To simulate CR models Cognitive radio network (CRN) simulators are used. In this paper we have investigated various open source CRN simulators. We have chosen two open source CRN simulators i.e. crSimulator and CRE-NS3. To study, evaluate and compare the capabilities we have analyzed the core features e.g. Supported CR Media Access Control (MAC) protocols, Mobility models etc. and parameters such as the ability to simulate large CRN, maximum number of PUs etc. Key parameters in the performance metric to compare network simulators are memory consumption, CPU utilization and throughput. Therefore we have simulated common CRN scenarios in both simulators. Our analysis shows that if low memory consumption and low CPU utilization are the requirements than CRE-NS3 is a suitable CRN simulator. If high throughput and large CRN (more than 200 CR nodes) are the requirement then crSimulator is a suitable CRN simulator.