Data fusion implementation in sensor networks applied to health monitoring
Ano de defesa: | 2005 |
---|---|
Autor(a) principal: | |
Orientador(a): | |
Banca de defesa: | |
Tipo de documento: | Tese |
Tipo de acesso: | Acesso aberto |
Idioma: | por |
Instituição de defesa: |
Universidade Federal de Minas Gerais
UFMG |
Programa de Pós-Graduação: |
Não Informado pela instituição
|
Departamento: |
Não Informado pela instituição
|
País: |
Não Informado pela instituição
|
Palavras-chave em Português: | |
Link de acesso: | http://hdl.handle.net/1843/RVMR-6EANVP |
Resumo: | Recent developments in wireless networks and in miniaturization ofpowerful embedded devices have enabled the development of very smallcomputing systems that are available all the time. In the literature, this type of computation has been called ubiquitous computing. Several applications of ubiquitous computing (including ones that cover life threatening situations) require fault tolerance, resilience and graceful degradation in response to different types of failures in the system. Several authors have focused on the development of middleware solutions to ease the design of ubiquitous computing applications.Others have addressed the application development field, but very few authors have addressed the relationship between middleware and application development. Data fusion is an important component of applications for systems that use correlated data from multiple sources to determine the state of a system. The fault tolerance and resilience of these applications will depend greatly on the data fusion framework. As the state of the system being monitored and available resources change,the general data fusion framework should change dynamically based on the current environment and available resources in the system. As a consequence, a general data fusion framework should provide some results of the data fusion to a module called the decision system. This module is responsible for sending feedback to the middleware, so the middleware can appropriately reconfigure the network. Based on the current data or variable, the decision system receives from the data fusion module, the decision system should automatically inform the middleware of the applications new requirements (i.e., the application dynamically adjusts its Quality of Service requirements based on the current state of the system being monitored and informs the middleware of its current Quality of Service needs). In this thesis we address the problem of how to implement data fusion in sensor networks, taking into account fault tolerance, resilience, and graceful degradation in a ubiquitous computing environment. We think that to achieve thesegoals it is necessary to develop applications upon a dynamic data fusionarchitecture. To achieve this goal we have created a new data fusion architecture; developed a software infrastructure based on this architecture and applied to centralized and distributed implementations; and developed a communication approach between middleware and the application. All these tools are new in the literature and represent important contributions to the data fusion implementation in sensor networks field. Furthermore, the combination of these tools represent animportant contribution to sensor networks applications development.As a proof of concept, we have developed a Personal Multi Parametric Heart Rate Monitor application based on sensor networks conception. The Personal Heart Rate Monitor consists of a body-worn sensor networks application powered by battery and connected by a wireless network. Therefore, resources such as channel bandwidth and node energy are limited, and must be managed efficiently. The developed system is based on the proposed tools to implement data fusion in sensor networks applications which dynamically adapts as the state of the persons vital signals change and provide graceful degradation to resource changes. The Personal Multi Parametric Heart Rate Monitor developed demonstrated to be an important contribution to the medical field. It will be tested in a clinical trial to evaluate its impact in prevention and early diagnosis of diseases. |