A survey of heuristics for matrix bandwidth reduction

Abstract

This paper surveys heuristic methods for matrix bandwidth reduction, including low-cost methods and metaheuristics. This optimization min–max problem represents a demanding problem for heuristic methods. This paper poses the graph layout problem with its formal definition. The study also considers the application domains in which practitioners employ the linear graph layout problem on general matrices. Furthermore, this paper focuses on the techniques and procedures that provide excellent results and provides an extensive perspective of approaches to devise heuristics for matrix bandwidth reduction. Thus, this paper surveys the most significant research in the field and examines the current state-of-the-art heuristics for the bandwidth reduction problem.