Microphone arrays have shown substantial capability in re- ducing noise due to their spatial selectivity. In this paper, we first introduce two noise reduction algorithms using mi- crophone arrays to deal with both coherent and incoherent noise components. Two algorithms have the same compo- nents, a generalized sidelobe canceller(GSC) and a Wiener filter, but with different structures. Their theoretical perfor- mance are then analyzed in different theoretically defined noise fields. The performance is further evaluated by using the real-world multi-channel noise recordings. With regard to the theoretical and experimental results, some discussions on two algorithms are also presented.