Because of the UWB(Ultrawideband) channel multipath sparsity, its estimation is based on sparse assumption. It usually adopts the compression channel sensing technology. However, in the indoor environment, scattering which is caused by multi obstacle environment often leads to ultra wideband channel presents a sparse and dense hybrid mode. Therefore, in consideration of the scattering characteristics of the case, based on the STOMP (stagewise orthogonal matching pursuit) compressed sensing algorithm, this paper proposes a novel channel estimation method which called STOMPMMSE. The algorithm overcomes the assumption based on sparse which ignore the scattering component of the channel, improving the accuracy of channel estimation. The simulation results show that the method proposed can help ultra wideband hybrid channel which contains scattering component to obtain the estimation error smaller.