We assume network trees consisting of mobile, energy constrained and rechargeable nodes as well as a static sink which collects the monitoring data and it is the root of the tree. Almost exhausted nodes can autonomously move towards a charging point to recharge their battery. However, this action leads to network disconnections and reduced lifetime since one or more predecessor nodes cannot forward their data to the sink. To alleviate this problem and extend network lifetime we examine the feasibility of replacing almost exhausted nodes using nodes with higher remaining energy. Based on this idea we propose a localized algorithm to autonomously replace nodes with high communication burden by the leaves of the tree. Both theoretical and simulation results show a big improvement in terms of network lifetime extension compared to the case where no replacement is performed and to the case where rerouting is considered.
~100 nodes, 10K sq.m. terrain size, CDS clustering, no recharging (file)
|Simple method. No replacement/rerouting.||Rerouting is performed in case of a disconnection.||Replacement is done using another node in the network.|
|Normalised network lifetime: 91.75 time units.||Normalised network lifetime: 260.51 time units.||Normalised network lifetime: 312.10 time units.|
Videos can be viewed using the following browsers (or their newer versions):
Mozilla Firefox 3.5, Chromium r18297, Google Chrome 3.0, Opera 10.50