Investigación Científica

El laboratorio CPU cuenta con diferentes áreas de Investigación científica....

Azcatl: A Bio-inspired data foraging mechanism based on indirect communication in sparse mobile sensor networks for dynamic environments

Sparse Wireless Sensor Networks (SWSN) have been proposed to explore and gather data from harsh environments. Exploration has been focused on finding optimal paths to obtain knowledge about the environment while data gathering has mainly been oriented to obtain raw data indiscriminately of the environment. However, dynamic environments were the usefulness of data is constantly changing, these solutions are not useful. Therefore, for these types of environments, both exploration and data gathering must be focused on repeatedly searching for valuable data sources over a particular area; as some animals perform foraging. Data foraging is a continuous process which involves: exploration, identification and selective exploitation (harvesting) of profitable data sources. We propose a bio-inspired data foraging mechanism for SWSN in dynamic environments. To develop our mechanism, indirect communication is modeled taking inspiration from the stigmergy principe. We establish artificial pheromones based on the fuzzy causal relation. Each step of data foraging is formally defined and mathematically evaluated. In addition we show the viability of each step by simulations which have been tested through statistical hypothesis.

Autor: Josué Castañeda Cisneros

Asesores
Dr. Saúl Eduardo Pomares Hernández (INAOE)
Dr. Jesús Antonio González Bernal (INAOE)
Dr. José Roberto Pérez Cruz (UMSNH)

Indirect spatio-temporal communication for SMWSN-based collaborative data-foraging in dynamic environments

In a Sparse Mobile Wireless Sensor Network (SMWSN) the mobile sensors, called Mobile Data Collectors (MDCs), are spread over a large area to retrieve environmental data. Due to the dynamic movement of the MDCs, they should not be restricted to share the same area at the same time. It is sometimes impossible to say that one of two samples occurred first, especially if the samples were obtained by different MDCs due to the lack of perfectly synchronized clocks. In this kind of conditions, to exchange information in a direct manner is not possible without an enduring transmission link. However, for some applications, it is required for the MDCs to collaborate and exchange information about constantly changing features. To achieve a profitable collaboration among MDCs, an essential task is to determine the context of the retrieved data; which requires relating the spatial and temporal domains in a feasible way. We propose an indirect spatio-temporal communication among a group of MDCs oriented to retrieve data applying a collaborative approach for dynamic environments. This approach is inspired from stigmergy to accomplish an indirect communication among MDCs to exchange information. The indirect spatio-temporal communication is performed through spatial-temporal relations based on causal and fuzzy causal dependencies.

Autor: Fabiola Marcos Solis
Asesores
Dr. Saúl Eduardo Pomares Hernández, INAOE
Dra. Lil María Xibai Rodríguez Henríquez, CONACYT-INAOE
Dr. José Roberto Pérez Cruz, CONACYT-UMSNH

Mecanismo escalable para la sincronización de datos multimedia en tiempo real para la comunicación en grupo

La necesidad de interacción entre varios usuarios ha motivado el desarrollo de nuevas aplicaciones en Sistemas Multimedia (SM), mediante estas nuevas aplicaciones se espera maximizar la experiencia del usuario y cumplir con la demanda de comunicación. Estos nuevos retos motivan el diseño y desarrollo de mecanismos para la sincronización de datos multimedia en Sistemas de Comunicación en Grupo (SCG). El principal problema al que se afrontan estos nuevos mecanismos es la escalabilidad, ya que deben de manejar de manera eficiente el incremento tanto en el número de flujos como de participantes en la comunicación. La principal contribución de esta investigación es la propuesta de un mecanismo escalable para la sincronización de flujos multimedia basado en un esquema descentralizado. El esquema propone la división del grupo de comunicación en subgrupos de participantes y una jerarquía en función de su distancia temporal. Mediante el uso de un esquema maestro/esclavo se logra jerarquizar los flujos; además se logra reducir el número de información de control a enviar y brindar el soporte al incremento de las entidades a partir de la división de subgrupos.

Autor: José Ricardo Pérez Cruz
Asesores
Dr. Gustavo Rodríguez Gómez, INAOE
Dr. Saúl Eduardo Pomares Hernández, INAOE
Dr. Eduardo López Domínguez, LANIA

Mecanismo de reconocimiento grupal del contexto orientado a los sistemas inteligentes de transporte

Esta tesis propone un mecanismo para el reconocimiento de las condiciones de los caminos por los cuales un vehículo se desplaza, como parte de la información contextual (o información del entorno) de los conductores. El reconocimiento de la información del contexto se realiza desde el enfoque de sensado móvil. Esto significa que se hace uso de sensores en dispositivos móviles para recolectar la información del entorno del conductor; específicamente en este trabajo se utilizaron acelerómetros de varios teléfonos inteligentes. Debido a su amplia usabilidad en varias aplicaciones de los sistemas inteligentes de transporte, en este trabajo se consideró relevante el reconocimiento de la información del contexto referente a las siguientes clases: 1) calles en buen estado, 2) calles en mal estado, 3) calles en estado regular, 4) topes, 5) baches, y 6) el estado de reposo. La aportación principal de esta tesis radica en conseguir un reconocimiento de información del contexto de forma eficiente, a través de un sensado grupal. Principalmente, el mecanismo propuesto consiguió mejorar la eficiencia en el uso de los siguientes recursos del teléfono: memoria, almacenamiento, procesamiento y batería. La mejora en la eficiencia se consiguió reduciendo la frecuencia de muestreo de cada dispositivo durante el sensado y realizando una fusión de datos centralizada. Los experimentos realizados mostraron que usando cuatro teléfonos con una frecuencia de muestreo de 25 Hz, se consiguen resultados comparables a utilizar la frecuencia máxima de muestreo permitida con una de las fuentes sola. Dado este comportamiento, es posible concluir que el mecanismo propuesto, basado en un sensado grupal, consigue un ahorro de recursos y es eficiente respecto al caso de sensado individual.

Autor: Miguel Ángel Valencia Serrano
Asesores
Dra. María del Pilar Gómez Gil
Dr. Saúl Eduardo Pomares Hernández, INAOE

Network Consistency Mechanism for Software-Defined Networks based on Causal Principles

In the last decades, computer networks have been evolved significantly. When network size increases, monitoring and troubleshooting will be necessary. However, existing tools of current network do not meet needs of administrators due to the complexity and the highly asynchronous and distributed environments. In the recent few years, a new network paradigm called Software-Defined Network (SDN) was proposed. The key idea of SDN is to separate the network control logic from data forwarding operations to create, independently from the forwarding devices, a centralized control platform for supporting a direct and flexible network management. However, SDN remains inherently distributed and asynchronous. This can lead to break network invariants such as, forwarding loop free, forwarding blackhole free, waypointing, etc. which referred as network inconsistency.

In this research, we are interested to ensure network consistency in SDN to enforce the correctness of data plan when network traffic evolves and configuration policies change. As SDN is an event-driven system, then a large space of event ordering should be considered. Network events can be executed asynchronously at any time from any network entity. This can cause an out-of order execution of events which can give rise to non-deterministic behavior. This leads to the main following problem: the network-wide view, from the centralized control entity (control plan), will be inconsistent with the network actual state. This inconsistent view could cause network behavior to deviate from invariants temporarily or even permanently. We claim that, when network policies change and network traffic evolves, network consistency can be achieved by establishing causal dependencies. Under such principles, we aim to design and develop a mechanism to ensure fundamental correctness properties against invariant violations that can occur, which serves as support to achieve network consistency in SDNs.

Autor: Amine Guidara
Asesores
Dr. Saúl Eduardo Pomares Hernández, INAOE
Dra. Lil María Xibai Rodríguez Henríquez, CONACYT-INAOE

Dependability for ESB systems in critical environments based on self-healing and checkpointing principles

Ensuring dependability for computer systems based on fault tolerance is an open challenge. Due to the complexity and heterogeneity of the interactions and services, offered by distributed systems, they make the administration and management of resources in highly dynamic environments exceed the capabilities of more experienced network administrators. As a consequence, new paradigms emerge (autonomic computing) and the Service-Oriented Architecture (SOA). In this dissertation, on the one hand, autonomic computing is focused on solving the complexity of monitoring and diagnosing the behavior of the systems with low resources and little human intervention. On the other hand, to consider the heterogeneity of the systems, the SOA paradigm and the Enterprise Service Bus (ESB) were used for their integration. Applications and systems are needed to intercommunicate with each other, often in unreliable environments such as the nature of the Internet. There are sophisticated solutions, such as replication of services, rollback recovery, and self-healing, which increase the systems' reliability. However, these approaches have drawbacks; for example, they affect the performance of the system, have high implementation costs and/or can endanger its scalability. On the other hand, to facilitate the self-management of the systems, in this dissertation, we implement the Monitoring, Analysis, Planning and Execution (MAPE) control cycle. An open challenge for the MAPE cycle is to efficiently carry out the diagnostic and decision-making processes, collecting data from which the system can detect, diagnose and repair potential problems, that is, increase the dependability of the systems specifically with fault tolerant mechanisms. A useful tool for this purpose is through the implementation of communication-induced checkpointing (CiC) mechanism. The experimental results support the viability of our proposals.

Autor: Mariano Vargas Santiago
Asesores
Dr. Saúl Eduardo Pomares Hernández
Dr. Luis Alberto Morales Rosales

Bio-Inspired Fuzzy-Causal Communication Protocol for Vehicular Ad-Hoc Networks

Emerging applications based on vehicular ad-hoc networks (VANETs) are focused on offering new services to Intelligent Transportation Systems, such as collective traffic monitoring. These applications require that vehicles and road side units exchange information among them. However, the constraints of VANETs, such as rapid topology changes, variable traffic flows/densities and dynamic behavior of vehicles make traditional solutions inefficient. Despite of these constraints, VANETs have important advantages such as bounded movements of cars throughout a road infrastructure. This research work presents the development of a bio-inspired communication protocol for VANETs. The proposed protocol disseminates the information in a specific geographical region among the fixed and mobile elements of a VANET, without a dedicated global communication infrastructure nor saturating the communication links. Furthermore, the protocol performs a spatial-temporal ordering of messages to ensure the system’s consistency. The proposed solution can be considered scalable, as the size of the control information does not depend on the number of entities in a system and does not incur any significant increment in system implementation and maintenance costs.

Autor: Grigory Evropeytsev
Asesores
Dr. Saúl Eduardo Pomares Hernández, INAOE
Dra. Lil María Xibai Rodríguez Henríquez, CONACYT-INAOE
Dr. José Roberto Pérez Cruz, CONACYT-UMSNH