Abstract. A kernel technology is proposed for a wide class of discrete optimization problems. Based on the notion of kernel, the technology implements stochastic perturbations for the iterative algorithmic schemes. Computational study of the proposed technology for the quadratic assignement problem demonstrated efficiency of this technology in terms of speed and solution quality. The kernel method can be easily incorporated into the available algorithms.

Keywords: kernel technology, QAP, WMaxCut, UBQP, TSP, JSP, computing experiment, a comparative study of algorithms.