Institution profile

University of Malaga

Academic institutioneurope · es
Official website
Research library110linked papers
Opportunities0open roles
Selected work

Representative Papers

Fast energy-aware OLSR routing in VANETs by means of a parallel evolutionary algorithm

Apr 27, 2012Cluster Computing

To address the high energy consumption and slow convergence of the Optimized Link State Routing (OLSR) protocol in vehicular ad hoc networks (VANETs), this paper proposes an energy-aware routing optimization method based on a parallel evolutionary algorithm. Specifically, a parallel genetic algorithm is integrated into the OLSR control plane, coupled with distributed fitness evaluation and a mobility prediction model to jointly optimize topology awareness, energy balancing, and low-latency path selection. Furthermore, the OLSR protocol is extended to support dynamic feedback of node energy states. NS-2 simulation results demonstrate that, compared to standard OLSR, the proposed approach reduces average energy consumption by 32%, accelerates routing convergence by a factor of 2.1, and decreases end-to-end delay by 27%. These improvements significantly enhance both energy efficiency and real-time performance in VANETs.

62 citations4 influentialRead paper

Automatic tuning of communication protocols for vehicular ad hoc networks using metaheuristics

Aug 01, 2010Engineering applications of artificial intelligence

To address the challenges of complex parameter configuration and poor dynamic adaptability in vehicular ad hoc network (VANET) communication protocols, this paper proposes a cross-layer cooperative adaptive tuning framework based on a hybrid metaheuristic algorithm. For the first time, genetic algorithm (GA) and particle swarm optimization (PSO) are synergistically integrated to jointly optimize critical parameters at both the MAC and routing layers. The framework is implemented and evaluated in the NS-2 simulator using an IEEE 802.11p-based VANET model. Experimental results under high-mobility scenarios demonstrate significant performance improvements: end-to-end packet delivery ratio increases by 32%, average throughput rises by 27%, and end-to-end delay decreases by 41%. These results validate the framework’s robustness and effectiveness in dynamically varying channel conditions and network topologies.

54 citationsRead paper

Light commodity devices for building vehicular ad hoc networks: An experimental study

Feb 01, 2016Ad hoc networks

To bridge the performance gap between simulation-based evaluation and real-world VANET deployments, this paper presents the first systematic validation of low-cost, off-the-shelf hardware—specifically Raspberry Pi paired with USB Wi-Fi adapters—for supporting core VANET functionalities in realistic vehicular mobility scenarios. Leveraging a Linux-based embedded platform, we implement IEEE 802.11p channel bonding, a customized MAC-layer simulator, the OLSR routing protocol, and joint GPS/IMU-based motion modeling. Experimental results under urban driving conditions (30–60 km/h) demonstrate an end-to-end message delivery ratio of 78% and an average latency below 1.2 seconds. This work challenges the conventional reliance on proprietary onboard units (OBUs), empirically confirming that lightweight commercial hardware can enable practical, small-to-medium-scale urban VANET deployments. It thus establishes a novel, cost-effective, and scalable paradigm for empirical VANET research and field experimentation.

27 citations1 influentialRead paper

Performance analysis of optimized VANET protocols in real world tests

Jul 04, 2011International Wireless Communications & Mobile Computing Conference

Existing studies lack empirical validation of vehicular ad hoc network (VANET) protocols—such as the Vehicular Data Transport Protocol (VDTP)—under realistic road-testing conditions. Method: This paper establishes a city-scale, real-vehicle-based VANET testbed and, for the first time, applies five metaheuristic optimization algorithms—Particle Swarm Optimization (PSO), Differential Evolution (DE), Genetic Algorithm (GA), Evolution Strategy (ES), and Simulated Annealing (SA)—to automatically optimize and deploy VDTP parameters on physical vehicles. Contribution/Results: The optimized VDTP achieves significant improvements over manually configured baselines: +23.6% average data delivery success rate and −31.4% average end-to-end latency. Field measurements closely align with simulation results, validating simulation fidelity. This work bridges a critical gap in systematic, real-world evaluation of VDTP and provides a reproducible methodology and empirical foundation for adaptive protocol parameter optimization in intelligent transportation systems.

24 citations1 influentialRead paper

Infrastructure Deployment in Vehicular Communication Networks Using a Parallel Multiobjective Evolutionary Algorithm

Aug 01, 2017International Journal of Intelligent Systems

This study addresses the roadside unit (RSU) placement problem for vehicle-infrastructure cooperative networks in realistic urban road environments, aiming to jointly optimize service quality (QoS) and deployment cost. We propose a novel multi-objective optimization framework that integrates traffic flow modeling, GIS-based spatial constraints, and multimodal communication load simulation (text, audio, video). For the first time, a parallel multi-objective evolutionary algorithm (PMOEA) is applied to this domain. Evaluated on the real-world map of Málaga, Spain, our approach efficiently generates high-precision Pareto-optimal solution sets. Compared with state-of-the-art methods, it significantly improves coverage quality and connection reliability while reducing RSU deployment costs by 12.7%–18.3%. The framework provides a scalable, reproducible, and intelligent optimization solution for city-scale RSU planning.

23 citationsRead paper
Recent publications

Latest Papers

The energetic cost of mitigating AI attacks in cellular networks

Aug 12, 2026

This study addresses the critical yet overlooked energy overhead incurred by adversarial defense mechanisms in AI-driven cellular networks. While existing defenses enhance model robustness against adversarial attacks, they introduce substantial energy costs that remain unquantified. This work presents the first systematic characterization of the energy consumption of AI defense strategies within the O-RAN architecture. By simulating adversarial attacks and deploying representative defenses, the authors construct a trade-off model among accuracy, robustness, and energy efficiency. The findings reveal an inherent tension among security, performance, and energy efficiency, offering both theoretical insights and empirical evidence to guide the design of lightweight, energy-efficient AI security mechanisms for next-generation wireless systems.

0 citationsRead paper

A continually expandable foundation model for brain MRI

Aug 08, 2026

This work proposes Alcmaeon, a continuously extensible 3D foundation model for brain MRI that addresses the limited generalizability and catastrophic forgetting prevalent in existing models, which are often constrained to specific diseases, populations, or imaging protocols. Trained unsupervised on over 425,000 unlabeled brain MRI scans, Alcmaeon incorporates a novel Graph-Blueprint Pruning mechanism that dynamically preserves critical network modules when integrating new clinical domains, thereby mitigating catastrophic forgetting. The model demonstrates consistently lower voxel-level reconstruction forgetting across diverse applications—from healthy aging to tumor imaging—and its hierarchical representations effectively support multiple downstream tasks, including image synthesis, disease classification, survival modeling, and postoperative outcome prediction, while providing interpretable records of capacity allocation.

0 citationsRead paper