Search

Emerging applications in the area of wireless sensor networks that include a wide variety of scenarios typically involve a signiicant number of nodes deployed in a wide area. To ensure proper communication between the network nodes to the base station, the messages are propagated by intermediate nodes, so a route with multiple links is established. In general, sensor nodes are characterized by limited resources; therefore it is essential to use protocols that ensure not only communication but also that save the most energy while providing scalability. Different routing protocols have been proposed for WSNs, but the reactive routing algorithms have shown be more energy eficient, which makes them suitable for wireless sensor networks. In this paper the performance of the reactive type protocol (Ad-Hoc On Demand Distance Vector) is evaluated under the simulation platform OMNeT++ through different scenarios. The AODV protocol has been chosen by the performance superiority comparing with other reactive protocols reported in the literature. The protocol has been evaluated under the following metrics: packet delivery rate, average end-to-end delay, routing overhead and energy consumption. The results show that AODV reduces their performance when the number of nodes increases, consequently it restricts the network scalability, further results show that the spatial distribution of the nodes affect on the performance of the protocol, speciic in our work, the uniform distribution shows better results. [ABSTRACT FROM AUTHOR]

In presented article, it looks for a new form to solve to the problem of routing and allocation of wavelength in networks photons (RWA - Routing Wavelength Assignement). This problem has been solved dividing it in 2 subproblems, and applying to different algorithms optimizers, among them are the genetic algorithms, he himself who is used in this investigation, nevertheless has seted out the solution without dividing it, solving it of integral way. The investigation analyzes some parameters of the genetic algorithms, and they are optimized for the later use in the resolution of problem RWA. For it, simulations were made in scenes of dynamic traffic in network NSFNET. It was managed to determine the position of the reproduction, the size of reproduction and the loss of life, that optimizes the algorithm, in this work was not used mutation, due to its little relevance in improving the algorithm. Later simulations of the optimized algorithm were made in the same network, under dynamic traffic, these results will be shown in a next publication. [ABSTRACT FROM AUTHOR]