Cognitive Radio (CR) systems are smart systems capable of sensing the surrounding radio environment and adapting their operating parameters for efficient utilization of the available radio spectrum. To reach this goal, different operating parameters across the OSI layers of a device should be optimized such as transmit power, modulation scheme and packet length etc. In this paper, we propose an efficient algorithm for optimizing and adapting CR operating parameters from physical, MAC and network layers for better utilization of radio spectrum. We include Automatic Repeat reQuest (ARQ) inside our proposed algorithm to check the validity of best parameters at the receiver side under specific link state. The proposed algorithm is called ARQ-Adaptive Discrete Particle Swarm Optimization (ARQ-ADPSO). The simulation results show an improvement by ARQ-ADPSO compared to the well-known genetic algorithm.