Routing algorithms for cognitive radio networks should consider various objectives and constraints. Constraint-based routing (CBR) algorithms select a route satisfying constraints that are either administrative-oriented (policy routing), or service-oriented (QoS routing). In cognitive radio networks, a primary user's interference constraint is the most important objective for the activities of secondary users. Beside the interference constraints, transmission delay is another important constraint for delay-sensitive services. In this paper, we propose a CBR algorithm that considers interference and delay constraints for a cognitive radio network. While these constraints address the QoS requirements for a cognitive radio network, to also consider the routing policy in the proposed CBR, we use the Pareto-optimal concept to allow the secondary users to transmit based on the network policy and resource constraints. The evaluation results of the proposed CBR scheme show that it can outperform other recently proposed CBR approaches.
