Abstract: With the introduction of interference efficiency(IE),which represents the interference introduced by the transmission of each data bit,an optimal dynamic spectrum management scheme was proposed to deal with the problem of reducing both interference and power consumption in small-cell cognitive radio networks working on unlicensed spectrum bands.Both the multi-objective reward function with energy and interference efficiency and spectrum management decisions to maximizing their long-term discounted reward were established and optimized by the cognitive radio base stations in the small cells.The restless bandits problem based on the Markov decision process in the dynamic stochastic optimization theory was used for the optimization schemes,and the process of dynamic spectrum management was presented.The proposed scheme has low computation complexity with distributed and dynamic features.Simulation results are provided to demonstrate the significant performance improvement of the proposed scheme.