The exploration of one-factorizations of complete graphs is the foundation of some classical sports scheduling problems. One has to traverse the landscape of such one-factorizations by moving from one of those to a so-called neighbor one-factorization. This approach amounts to modifying locally the coloring associated with a one-factorization. We consider some particular types of modifications and describe various constructions which give one-factorizations which may be modified or not by these techniques. Among those are recoloring of bichromatic cycles, altering of optimally colored subcliques of even size, or recoloring of chordless lanterns.

In this paper we present a detailed experimental analysis of the existing neighborhoods in the literature using a few quality criteria and propose new heuristics for the Mirrored Traveling Tournament Problem. Based on the Home-Away Swap, the proposed heuristics are able to search the solution space from an infeasible solution and obtaining a feasible one in the end. We also describe a estrategy for reduce the amount of visited neighborhoods in the local search process without drastically affect the solution, consequently we managed to obtain feasible solutions with reduced computation time, as shown through empirical analysis.

Wireless Sensor Networks (WSNs) are composed of autonomous and resource-constrained (power, sensing, radios, and processors) nodes. These networks are conceived to have a large number of nodes working on monitoring phenomena. A major challenge for theses networks is to provide solutions that maximize quality of service (QoS) requirements, such as coverage and data routing, and minimize the energy consumption. This paper presents an evolutionary algorithm (EA) to solve the Density Control, Coverage and Routing Multi-Period Problem (DCCRMP) in WSN. The results are compared to the optimal solutions obtained by an Integer Linear Program model and a GRASP heuristic from the literature. The EA obtains significant improvements in both quality of solutions and computational time.

This article considers the bi-objective Job Shop Scheduling Problem in which the make span and the total tardiness of jobs are minimized. In order to find a set of dominant solutions, that is, an approximation of the Pareto optimal solutions, we propose three versions of a genetic algorithm with techniques like hybridization with local search, path relinking and elitism. The three versions of the algorithm are compared with each other and they are also compared with other multiobjective genetic algorithm proposed in the literature.