Grupo de Analisis

Spain

Grupo de Analisis

Spain

Time filter

Source Type

Villalba L.J.G.,Grupo de Analisis | Villalba L.J.G.,Complutense University of Madrid | Canas D.R.,Grupo de Analisis | Canas D.R.,Complutense University of Madrid | And 2 more authors.
IET Communications | Year: 2010

A mobile ad hoc network (MANET) is a non-centralised, multihop, wireless network that lacks a common infrastructure. It therefore needs self-organisation. The MANETs are autonomous, adaptive and dynamic. In MANETs, multihop routing allows multiple routes between the source and destination to be established. This study speaks about a bio-inspired routing protocol for these networks based on AntHocNet. The design for the protocol lies in a heuristic, based on swarm intelligence, which takes into account the limited resources and highly dynamic environment, as well as the restriction on the exchange of routing information. So, the key aspects of the proposed protocol are the disjoint-link and disjoint-node routes, separation between the regular pheromone and the virtual pheromone in the diffusion process and the exploration of routes, taking into consideration the number of hops in the best routes which the authors have previously found out. The simulation results show that the protocol has lower overhead and higher delivered packet ratio than AntHocNet. Likewise, these results indicate that the routing satisfies multiple and independent quality of service constraints and can deal with faults, which provides better load balancing. © 2010 The Institution of Engineering and Technology.


Villalba L.J.G.,Grupo de Analisis | Canas D.R.,Grupo de Analisis | Orozco A.L.S.,Grupo de Analisis
International Journal of Ad Hoc and Ubiquitous Computing | Year: 2013

Designing routing protocols for Mobile Ad Hoc Networks (MANETs) is a complex task because of its dynamic topology. A kind of routing protocols that suits the particularity of MANETs is so-called bio-inspired. Among these, focused on Ant Colony Optimisation (ACO), which studies the behaviour of ants in their search for food, are especially relevant. One of these algorithms is AntOR, which relying on swarm intelligence, efficiently solves routing in MANETs. In this paper we show a parallelised version of AntOR, the so-called P-AntOR, that using programming multiprocessor architectures based on shared memory protocol, allows to run tasks in parallel using threads, being applicable this parallelisation in the route discovery phase, route local repair process and link failure notification. The simulation results indicate that P-AntOR performs better than its predecessor, with emphasis on the metric of average End-To-End delay, jitter and packet delivery ratio. © 2013 Inderscience Enterprises Ltd.


Canas D.R.,Grupo de Analisis | Orozco A.L.S.,Grupo de Analisis | Villalba L.J.G.,Grupo de Analisis
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2012

Ad hoc networks are formed by wireless devices distributed without a predefined infrastructure using a technique called multi-hop communication. A particular case is mobile ad hoc networks, which operate within dynamic environments. This determines the necessity of paying special attention to the routing problem. Traditional techniques are not particulary efficient at making the bioinspired algorithms more relevant. These techniques are based on the analysis of the behavior of some animals, especially in the process of obtaining food. A set of these techniques or algorithms are known as the ACO (Ant Colony Optimization) which is based on the particular behavior of ants. A representative protocol from this kind is AntOR, routing protocol for mobile ad hoc hybrid, multipath and adaptive. In this article a variant of AntOR is proposed which causes the protocol AntOR-UDLR. This approach consists of replacing the link failure notification messages sent in a broadcast manner by unicast messages, which are sent to the predecessor of the node reporting the link failure, until the source of the data session is reached. The simulation results show that AntOR-UDLR improves its predecessor according to all analyzed metrics. © Springer-Verlag Berlin Heidelberg 2012.

Loading Grupo de Analisis collaborators
Loading Grupo de Analisis collaborators