Strong global convergence properties of algorithms for nonlinear symmetric cone programming

Sequential optimality conditions have played a major role in establishing strong global convergence properties of numerical algorithms for many classes of optimization problems. In particular, the way complementarity is handled defines different optimality conditions and is fundamental to achieving a strong condition. Typically, one uses the inner product structure to measure complementarity, which provides a general approach to conic optimization problems, even in the infinite-dimensional case. In this paper we exploit the Jordan algebraic structure of symmetric cones to measure complementarity, resulting in a stronger sequential optimality condition related to the well-known complementary approximate Karush-Kuhn-Tucker conditions in standard nonlinear programming. Our results improve some known results in the setting of semidefinite programming and second-order cone programming in a unified framework. In particular, we obtain global convergence that are stronger than those known for augmented Lagrangian and interior point methods for general symmetric cones.