This　paper　proposes　a　novel　direction　of　arrival　(DOA)　estimation　algorithm　that　can　rapidly　estimate　the　DOAs　of　incidence　signals.　The　conventional　multiple　signal　classification　(MUSIC)　algorithm　should　perform　an　eigenvalue-decomposition　which　has　high　computational　complexity.　In　addition,　the　MUSIC　algorithm　can＇t　be　used　under　complex　noise　environment.　In　contrast,　the　novel　algorithm　can　rapidly　obtain　the　signal　subspace　from　a　submatrix　of　a　contracted　quadricovariance　fourth-order　cumulants　matrix　and　avoid　eigenvalue　decomposition.　Computer　simulation　results　show　that　the　method　is　effective　and　still　has　good　performance　under　colored　Gaussian　noise,　non-Gaussian　noise　with　symmetrical　distributing　and　white　noise　because　of　the　use　of　fourth-order　cumulants.　©　Springer-Verlag　Berlin　Heidelberg　2012.