This proposal will investigate a general framework for signal recovery problems, combining set-theoretic and optimization techniques for use with incomplete measurement data and partial constraints. Computationally attractive algorithms and performance bounds will be developed and assessed, with the results applied to inverse problems of practical interest.