Abstract: How to make efficient spectrum allocation of cognitive wireless network is the key technology for dynamic spectrum access．This paper presented an improved binary bacterial foraging optimization algorithm with quantum variation operation based on the graph coloring theory model of spectrum assignment,and used the maximum system efficiency of cognitive wireless network as the objective function,achieving the free radio frequency spectrum’s dynamic allocation among the cognitive users．Simulations were conducted to compare this algorithm with color sensitive graph co-loring algorithm and traditional binary bacterial foraging optimization algorithm．Results show that the proposed algorithm has better performances．It can achieve the maximization of network utility and increase the second user’s average utility．Compared with the traditional binary bacterial foraging optimization algorithm,it has better optimization ability and faster convergence speed.

Key words: Cognitive wireless network,Spectrum assignment,Bacterial foraging optimization algorithm,Graph coloring,Quantum variation operation