Abstract

Ubiquitous computing systems which include wireless devices in their networks rely on sometimes fragile ah-hoc communication channels between heterogeneous devices to operate. To maintain stability and robustness, the contextual information contained on each node present in the system needs to be disseminated between nodes, so that in the event of a node failure, the data is still available to the wider network. One approach to disseminating information around topologically unstable networks is a type of epidemic algorithm called gossiping. However, developing an efficient, resilient algorithm to operate effectively in such an environment is an ongoing challenge. We have developed a visualisation tool that supports the evaluation of gossiping methods by enabling the algorithm designer to view the visual evolution of a network of nodes, from which they can monitor the stability of information within the network. By using this tool the designer can watch as information is propagated throughout their network, and see the effects of events such as node failures, so that they can make informed changes to the design of their algorithms.

Citations