In wireless system where multiuser detection techniques may be applied, the ambient channel noise is known through experimental measurements to be decidedly non-Gaussian. This paper consider the problem of adaptive robust detection of direct-sequence spread-spectrum(DS-SS) signal in the presence of unknown correlated SS interference and additive non-Gaussian noise. Many efforts have been made and most of them have two major problems:the first, explicit a priori knowledge of the underlying noise statistics is generally not available in a realistic scenario; second, computational complexity must be taken into consideration. From traditional subspace tracking techniques, we derive a robust method with low computational complexity. Analytical and simulation results show that the proposed robust techniques offer significant performance better than the normal subspace tracking methods in impulsive noise.