由于光纤光敏性的限制，当信道数N变得很大时，采用幅度取样方式制作多信道光栅将变得非常困难。通过相位取样来制作多信道光栅被证明是一种能有效降低峰值折变量的方法。提出一种采用逆散射技术和优化策略的混杂算法有效地设计多信道光栅。由遗传算法进行优化，得到每一个信道相应的相位因子的最优值，改进了多信道光栅复杂的反射谱。再由Layer－peeling算法为依据引入了相位因子后的频谱响应重构光栅。通过对信道间相对相位的优化，制作这种多信道光栅所需的峰值折变量降低为单信道光栅的√N倍。经过数值仿真，得到8信道光栅的峰值折变量大约为单信道光栅的√8倍。

With the limit of photo-sensitization,when the channel number becomes large,it's unrealistic to fabricate such sampled fiber Bragg gratings(FBG). Phase-only sampled FBGs were demonstrated to effectively reduce the maximum refractive index modulation(MRIM). A hybrid method which employs the inverse scattering(IS)technique and the optimization strategy is demonstrated to be efficient for the design of multi-channel fiber Bragg gratings. Optimized by genetic algorithm(GA),a set of optimal constant phases are introduced to the phase responses of the multiple channels,which lead to a modified complex reflection spectrum of the multi-channel fiber Bragg grating. Then,the layer-peeling IS algorithm is applied to extract the grating index profile from the modified spectral response. By introducing a set of optimal phases to the phase response of the multiple channels,the target reflection response of the multi-channel FBG is modified,which ensures the MRIM of the N channel grating is about only √N times that of the single channel grating. The simulation results show that this hybrid method reduces the MRIM required for 8 channels to approximate √8 times that of the single channel.