Back to overview

Extending Sensor Network Lifetime with Middleware that Actively Manipulates Network Parameters

English title Extending Sensor Network Lifetime with Middleware that Actively Manipulates Network Parameters
Applicant Murphy Amy
Number 108079
Funding scheme Project funding (Div. I-III)
Research institution Istituto Media e Giornalismo Facoltà di Scienze della comunicazione Università della Svizzera italiana
Institution of higher education Università della Svizzera italiana - USI
Main discipline Information Technology
Start/End 01.10.2005 - 30.09.2006
Approved amount 82'522.00
Show all

Keywords (3)

Sensor networks; middleware; energy management

Lay Summary (English)

Lay summary
Wireless sensor networks are growing in popularity due to theavailability of low-power sensing devices and their flexibility foruse in a variety of applications ranging from environmental monitoringto personal health management. Designing and building theseapplications is complex because of (1) the distributed nature of theenvironment, (2) the fact that the available sensors change over time,and (3) the nature of the applications themselves, which typicallyrequire strict quality of service (e.g., data accuracy) that maychange over time as the state of the system being monitored changes.Further, the application lifetime is often limited by the battery sizeof the sensors. Managing these energy reserves has been studied atthe protocol level, but this work often ignores the application needs.Recently attempts have been made to raise the abstraction from thenetwork into middleware for sensor networks. While this workaddresses energy concerns, it often provides only best effort to theapplication, without taking into consideration what the applicationactually requires. Consider a case where the application does notrequire the maximal information, but can instead function fully withpartial information. By limiting the sensors used at any given time,different sensors can provide information at different times, and thelifetime of the application can be increased.We have proposed a new middleware, MiLAN---Middleware LinkingApplications and Networks, that both addresses the needs of theapplication and reaches down into the network to address energyconcerns. The goal is to maximize the lifetime of the network whilemaintaining the quality required by the applications. Our previouswork defined an architecture for MiLAN, including an API forapplications to specify their needs and a plug-in structure for MiLANthat allows it to function over multiple network types (e.g.,Bluetooth, IEEE 802.11 and 802.15.4). With this framework in place,the work of this proposal is the research necessary for itsinstantiation.Our key idea is to allow the middleware to adjust the network specificparameters to manage energy consumption (e.g., selecting nodes asmaster/slave in Bluetooth, or specifying multi-hop routes for data802.11) but always in such a way that the application quality needsare met.This proposal is a continuation of a three year collaboration betweentwo teams of researchers, one at the University of Lugano, Switzerlandand the other at the University of Rochester, New York, USA. Separatefunding has been applied for in the USA in the form of an USA NationalScience Foundation CAREER proposal and other aspects are to beproposed in the scope of a Office of Naval Researchers YoungInvestigator's proposal (deadline: December 2004). The work of thetwo proposals, although complementary, are sufficiently independentthat if one part is not funded, the other can still continue. This isdue in part to the fact that the previous work laid a strongfoundation for the new research to build upon.The novelty of this project lies in the combination of sensor networktechnology and middleware. These two fields have developedindependently, and it is difficult for a single individual to have thedepth of knowledge of both fields to complete this research.Therefore, this proposal combines the effort of two PIs, one withbackground and experience in the design and development of sensornetwork protocols, and the other with the same skills in middlewarefor mobile computing applications. Based on our initial interactions,we believe that collaboration will lead to a balanced system thatmakes significant breakthroughs in understanding the key issues in thearea, identifying algorithmic solutions that can be applied inmiddleware besides our own, and possibly in other types of networks,such as mobile ad hoc networks. Generally, our work will facilitatebridging the gap in communication of ideas and research directionsbetween the sensor network and distributed middleware fields.
Direct link to Lay Summary Last update: 21.02.2013

Responsible applicant and co-applicants