Vortrag Diplomarbeit

Impact of Link Quality Estimation on Routing Protocols for Wireless Sensor Networks

Tian Tian

Wireless sensor networks have evolved rapidly in recent years. Nowadays, even application areas like industrial monitoring (example below) are targeted. These have been precluded for wireless sensors, since they have highest demands towards the reliability of network connections. Routing protocols which target this application area must provide a very high end-to-end reachability. Several methods are applied in state-of-the-art protocols to achieve this, e.g. channel hopping, multi- path routing, or retransmissions. A very interesting method is multi-path routing since it might enable the network to route packets around interferers and it can even cope with device failures. In the industrial environment there is heavy interference of the wireless channel in the 2.4 GHz ISM band, e.g. caused by machines, workers’ mobile phones, or microwaves in the break room. In order to handle these circumstances reliable routing protocols usually utilize information gained from an estimation of the link quality. But interferences are usually unpredictable and therefore the link quality is also changing unpredictably. Thus the estimation of the link quality might not always be correct. Algorithms used for estimation of link quality are based on measured parameters like packet reception rate or received signal strength.
Aim of this thesis is to examine how the accuracy of link quality estimation is influencing the performance of a routing protocol that relies on this information. Therefore, existing link quality estimator algorithms must be analyzed and a suitable model of a link quality estimator must be developed within an Omnet based simulation framework. Additionally, a routing protocol, e.g. WirelessHART (example above), must be implemented for the Omnet simulator which can be used for concluding evaluations.