DECONVOLUTION WITH BOUNDED UNCERTAINTY

In deconvolution problems there are two primary sources of uncertainty in the data formation mechanism, namely measurement noise and errors in the model of the system. In this paper we develop an abstract set t heoretic deconvolution framework for problems in which the only inform ation available about these sources of uncertainty consists of bounds. Iterative methods based on projections are used to generate solutions consistent with these bounds, the output data signal and a priori kno wledge about the input signal. An example of application of this gener al framework to discrete signal recovery is demonstrated.