为提升无线系统的能效，针对用户随时间的潮汐变化特点，文中提出了一种通过基站开关切换提升系统能效的方法——基于能效准则的基站开/关选择策略，将用户业务分为恒定比特率业务和可变比特率业务，在满足恒定比特率业务的约束下最大化能源效率． 文中采用改进的自适应遗传算法来求解此约束优化问题，即基站开关状态用二进制编码，通过设计的选择、交叉、变异实现迭代寻优． 仿真结果表明:文中提出的改进遗传算法收敛速度快，能有效对抗染色体群体的早熟问题;在满足恒定比特率用户需求的前提下，文中策略可有效地提升无线系统的能效，其性能接近穷举算法．

In order to improve the energy efficiency of wireless communication systems，by taking into consideration the tidal property of user demands，a strategy to improve energy efficiency for cellular system is proposed on the ba-sis of cell switch on-off． In this strategy，the demands of users are divided into constant-bit-rate (CBR) and varia-ble-bit-rate (VBR) demands to maximize the energy efficiency under the constraint that CBR demands are ful-filled． Moreover，an improved genetic algorithm (GA) is employed to solve the constrained optimization problem，and the switch on-off states of base stations are encoded with binary sequence，followed with specifically-designed selection，crossover and mutation operators． Simulated results show that the proposed GA algorithm achieves fast convergence and prevents prematurity of chromosome population; and that the proposed cell switch on-off strategy effectively improves system's energy efficiency and guarantees the demand of CBR service，so that it is of high per-formance close to the exhaustive searching algorithm．