The difficulty of maximizing the lifetime in directional sensor networks has gained increasing attention recently. Most of the existing studies are focused on directional sensors with single or several predefined sensing ranges. In the present study, directional sensors can change sensing ranges smoothly. We address the problem of maximizing the lifetime in directional sensor networks with such smoothly varying sensing ranges,and propose a hybrid approach that combines a column generation method with an immune genetic algorithm. We search for attractive columns with the genetic algorithm, and optimize them by designing dynamic vaccines. Computational results demonstrate the performance of the proposed approach. Meanwhile,the advantage of the mentioned sensors in terms of solution quality is also revealed. The difficulty of maximizing the lifetime in directional sensor networks has increased increasing attention recently. Most of the existing studies are focused on directional sensors with single or several predefined sensing ranges. problem of maximizing the lifetime in directional sensor networks with and smooth mutation sensing ranges, and propose a hybrid approach that combines a column generation method with an immune genetic algorithm. We search for attractive columns with the genetic algorithm, and optimize them by designing dynamic vaccines Computational results demonstrate the performance of the proposed approach. Meanwhile, the advantage of the mentioned sensors in terms of solution quality is also revealed.