Publication: Optimización de energía y eficiencia de transmisión con análisis de imparcialidad en comunicaciones inalámbricas adaptativas
Loading...
Identifiers
Publication date
2010-02
Defense date
2010-02-17
Authors
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
Las redes inalámbricas se caracterizan por utilizar un medio de transmisión dinámico, en el que las condiciones de transmisión cambian de forma continua. Este cambio continuado lleva a diseñar los sistemas de comunicaciones desde dos puntos de vista. El más elemental es diseñar nuestro sistema para el caso peor. Aunque sencilla, esta solución lleva aparejado un sobredimensionamiento de los recursos empleados; si consideramos por ejemplo la potencia de transmisión, este sobredimensionamiento resultará, entre las consecuencias más inmediatas, en mayores interferencias y mayor consumo de energía (y por tanto menor duración de las baterías en sistemas así alimentados). Frente al anterior enfoque, la transmisión adaptativa permite adecuar los parámetros de transmisión a las condiciones del canal. El objetivo de esta tesis es proponer esquemas de transmisión adaptativos para sistemas inalámbricos que, bien mejoran técnicas existentes, como veremos para redes de sensores, o bien resuelven problemas nuevos, como la optimización de la eficiencia de transmisión de una estación base. Contribuimos por tanto al desarrollo de técnicas de transmisión adaptativa en redes inalámbricas para conseguir sistemas eficientes en energía y en tasa de transmisión. Para ello, previamente realizamos un recorrido por las técnicas existentes de transmisión adaptativa (capítulo 2) y presentamos las herramientas matemáticas que se utilizarán para resolver los problemas de optimización que resultan de las técnicas propuestas (capítulo 3). La transmisión eficiente en energía que proponemos tiene como objetivo optimizar la energía total de transmisión del enlace ascendente para un sistema centralizado, con comunicaciones de salto único. Se asume que la red es heterogénea, lo que se traduce en que cada nodo tendrá unos requisitos diferentes de calidad de servicio en función de la aplicación que se ejecute en cada nodo. La calidad de servicio vendrá expresada mediante la tasa de error de símbolo y una tasa de transmisión de bit. La optimización propuesta se basa en un esquema de modulación cross-layer que asigna adaptativamente la duración de los intervalos de transmisión, y por tanto los bits por símbolo a transmitir, cumpliendo con la calidad de servicio exigida. Este esquema lo particularizamos para redes inalámbricas de sensores. La optimización de la eficiencia de transmisión (o goodput) para el enlace descendente entre una estación base y los usuarios que de ella dependen, la realizaremos mediante la asignación de los recursos disponibles: subcanal (entendido como agrupación de subportadoras), potencia y tasa de bit por símbolo. El modelo del sistema implementa un mecanismo de control de imparcialidad que evita que resulten beneficiados aquellos usuarios que presenten unas condiciones más favorables de transmisión. Para resolver la asignación de recursos, se proponen dos esquemas: un esquema de asignación en dos pasos que reduce la complejidad del problema de asignación inicial, y un esquema subóptimo, orientado a la implementación práctica, que permite una reducción muy considerable de la carga computacional, sin una degradación apreciable del rendimiento respecto al óptimo.------------------------------------------------------------------------------------
Wireless networks are characterized by the use of a dynamic transmission channel, in which transmission conditions are continuously changing. This continuous change asks for designing communication systems from two different viewpoints. The more basic one is to prepare our system for the worst case; this option, although straightforward, implies an excess of the resources used; for instance, if transmission power is considered, this excess of power results in a high interference and a high energy consumption (and, therefore, provokes a shorter lifetime for the batteries), among the most inmediate consequences. Contrary to the previous point of view, adaptive transmission allows to adjust transmission parameters values to channel conditions. The objective of this thesis is to propose new adaptive transmission schemes for wireless systems. Some of these schemes outperform existing techniques, as for the case of sensor networks, an others solve new problems, for instance the optimization of a base station transmitted goodput. We therefore contribute to the development of adaptive transmission techniques for wireless networks in order to achieve energy-efficient and transmision-efficent systems. To this end, we have previously reviewed the existing adaptive transmission techniques (chapter 2) and the mathematical tools that we apply to the optimization problems resulting from the proposed techniques (chapter 3). The energy-efficent scheme we propose targets the total transmission energy optimization in the uplink of a centralized system, with single-hop communications. Network heterogeneity is assumed, which translates into different quality of service requirements depending on the application running at each node. Quality of service is expressed by means of the symbol-error rate and the minimum bit-rate. The presented design is cross-layer-based and adapts time-slot duration to channel conditions, using adaptive modulation to fulfil the required quality of service. This proposal is particularly fitted to wireless sensor networks. Goodput optimization for the downlink is performed by means of resources allocation, specifically of subchannels, transmit power and bit-rate per symbol. For our purpose, a subchannel is defined as a group of subcarriers. Our system model is provided with a fairness control mechanism that prevents users with favourable channel conditions from obtaining most of the resources. We present a two-step assignment scheme to solve resource allocation that reduces the complexity of the original problem, and a suboptimal implementation-oriented approach that allows a significative reduction of the computational load, without a significant loss of performance.
Wireless networks are characterized by the use of a dynamic transmission channel, in which transmission conditions are continuously changing. This continuous change asks for designing communication systems from two different viewpoints. The more basic one is to prepare our system for the worst case; this option, although straightforward, implies an excess of the resources used; for instance, if transmission power is considered, this excess of power results in a high interference and a high energy consumption (and, therefore, provokes a shorter lifetime for the batteries), among the most inmediate consequences. Contrary to the previous point of view, adaptive transmission allows to adjust transmission parameters values to channel conditions. The objective of this thesis is to propose new adaptive transmission schemes for wireless systems. Some of these schemes outperform existing techniques, as for the case of sensor networks, an others solve new problems, for instance the optimization of a base station transmitted goodput. We therefore contribute to the development of adaptive transmission techniques for wireless networks in order to achieve energy-efficient and transmision-efficent systems. To this end, we have previously reviewed the existing adaptive transmission techniques (chapter 2) and the mathematical tools that we apply to the optimization problems resulting from the proposed techniques (chapter 3). The energy-efficent scheme we propose targets the total transmission energy optimization in the uplink of a centralized system, with single-hop communications. Network heterogeneity is assumed, which translates into different quality of service requirements depending on the application running at each node. Quality of service is expressed by means of the symbol-error rate and the minimum bit-rate. The presented design is cross-layer-based and adapts time-slot duration to channel conditions, using adaptive modulation to fulfil the required quality of service. This proposal is particularly fitted to wireless sensor networks. Goodput optimization for the downlink is performed by means of resources allocation, specifically of subchannels, transmit power and bit-rate per symbol. For our purpose, a subchannel is defined as a group of subcarriers. Our system model is provided with a fairness control mechanism that prevents users with favourable channel conditions from obtaining most of the resources. We present a two-step assignment scheme to solve resource allocation that reduces the complexity of the original problem, and a suboptimal implementation-oriented approach that allows a significative reduction of the computational load, without a significant loss of performance.
Description
Keywords
Comunicaciones inalámbricas adaptativas, Optimización energética