PTS技术是将输入的数据符号划分为若干个子组，每个子组的长度仍为N，然后对每个子组进行系数最优化的求解，最后再合并这些子组，从而达到降低整个系统的PAPR的目的。但是存在搜索最优相位向量集合时复杂度较高的问题，特别是子块增加时。迭代线性位移搜索法能降低复杂度但是性能太差，本文的分组搜索算法能在复杂度与性能之间折衷。-PTS technique is the input data symbols into several sub-groups, each sub-group length is still N, and then solving the optimization coefficients for each sub-group, and finally merge these sub-groups, which can reduce the PAPR of the entire system purposes. However, when searching for the presence of the optimal phase vector set higher complexity of the problem, in particular, the sub-block increases. Iterative linear displacement search method can reduce the complexity but poor performance, grouping search algorithm in this paper can compromise between complexity and performance.