ReSNA (Regularized Smoothing Newton Algorithm)
ReSNA is a free matlab software for solving Mixed Nonlinear Second-Order Cone Complementarity Problems (MNSOCCP).
By using the attached plugins, you can apply ReSNA not only to SOCCP but also many kinds of problems such as Linear Program (LP), Quadratic Program (QP), Nonlinear Program (NLP), Linear/Nonlinear Second-Order Cone Program (LSOCP/NSOCP), (Mixed) Linear/Nonlinear Complementarity Problem (Mixed LCP/NCP), (Mixed) Linear/Nonlinear Second-Order Cone Complementarity Problem (Mixed LSOCCP/NSOCCP), and Nonlinear Equation (NEQ). (Now, we only have plugins for complementarity problems.) Since ReSNA is based on the regularized smoothing Newton method, we can expect not only the global convergence but also the quadratic convergence. (The convergence is guaranteed theoretically under the assumption that the included function is (Cartesian) P0.)
 Shunsuke Hayashi, Nobuo Yamashita and Masao Fukushima, A combined smoothing and regularization method for monotone second-order cone complementarity problems, SIAM Journal on Optimization, 15 (2005), pp. 593-615. (DOI:10.1137/S1052623403421516)
 Shunsuke Hayashi, Guidance to ReSNA (in Japanese), Special Issue: Ride on a wave of second-order cone programming, Communications of the Operations Research Society of Japan, Vol.59, No.12, (2014), pp. 716-724. (download from ORSJ)
Back to Hayashi's Webpage