Efficient data harvesting in mobile sensor networks

Sensors are now attached to a lot of mobile objects such as animals and vehicles to collect data. This forms intermittently connected mobile sensor network due to weak communication power and nodal mobility, in which conventional routing protocols in wireless sensor network don´t work. In this paper, we propose a data harvesting scheme for intermittently connected mobile sensor network. This approach takes fully advantage of storage resource and mobility pattern knowledge to improve the delivery ratio while minimizing the transmission overhead. We evaluate the performance of the approach through the OMNet++ simulator with the realistic mobility traces. The simulation results show that the proposed scheme achieves high delivery ratio while controlling the transmission overhead. Furthermore, the robust approach can be adaptive to the dynamic topology of network.