TITLE

Maximizing the lifetime of battery-powered networks

PRESENTER

Maurits de Graaf, senior innovatie-engineer, Thales Huizen

ABSTRACT

Today’s wireless sensor networks and hand-held mobile devices support a variety of computation and communication functions, e.g. sensor networks to monitor temperature and CO₂ level in buildings or advanced PDA’s or professional wearable computing devices for police, firemen or other emergency services. The perceived quality of these devices depends heavily on their autonomy, which is, essentially, the capability to manage their limited battery capacity. The fact that these devices should be able to connect in an ad hoc fashion, with nodes relaying traffic for others, adds to the complexity of these issues.

In such networks there are two aspects to the concept system lifetime. One is the lifetime of the individual nodes in the network, which depends on the (local) workload and processing efficiency. The other aspect is the lifetime of the whole network. Here, the network lifetime is defined as the time until the first node runs out of energy. It is important to maximize the network lifetime: batteries are (relatively) heavy, large, and difficult (sometimes even impossible) to replace. In particular, the network lifetime depends on the assigned transmit powers and the chosen routing algorithm. This is because transmit power and route lengths (number of hops in a route) have a big impact on energy consumption.

There is a trade-off between the transmit power assignments and the lengths of routes. Assigning a high transmit power to the nodes, will result in short routes (small number of hops), hence in few nodes with relatively high energy consumption for a communications session. Assigning a low transmit power to the nodes, will result in long routes (big number of hops), hence in many nodes with energy consumption for a communications session.

In the presentation we discuss recent work in this area that has been carried out together with the University of Twente in the Casimir project: average analysis of a transmit power assignment algorithm, efficient selection of multipoint relay nodes in an OLSR-based network, and optimization of the network lifetime in a specific type of radio network.

BIO

Maurits de Graaf received a PhD degree at the University of Amsterdam in 1994 for the thesis ‘Graphs and curves on surfaces’. From 1994 to 1999 he was with KPN Research, working on traffic management methods for ATM networks. Since 1999, he has been with Thales Netherlands, working as a system engineer and since 2008 at the Innovation, Research & Technology department. He has been involved in many collaborative projects with other companies and research centres. The topics addressed in those projects mostly relate to the introduction of IP networking and wireless networking technologies in the Thales product portfolio. Examples are projects on IPV6, Voip, IEEE 802.11, IEEE 802.1q, and wireless routing algorithms. Since 2006 he was involved in the Casimir programme (funded by NWO), where he is working part-time at the University of Twente within the SOR (Stochastic Operations Research) group on the development and analysis of algorithms to increase the lifetime of wireless ad hoc and sensor networks.

LANGUAGES

Dutch