Robust array processing detectors in dependent noise

The theory of m-interval partitioning is extended and applied to the problem of sequentially detecting weak signals in time-dependent noise by using an array of sensors. Each sample vector of the data is transformed, via a suboptimum projection, into a single variable with univariate distribution upon which the sequential test is applied. The recursive evaluation of the projected data can reduce the complexity and the time to decision of the sequential detector considerably. The resultant array processor is easily implemented and is adaptable to slowly changing noise conditions.<>