Active data dissemination for mobile sink groups in wireless sensor networks

Abstract

In wireless sensor networks, a mobile sink group brings out many challenging issues with regard to data dissemination due to its twofold mobility: group mobility and individual one. All member sinks of a group should move together toward the same destination in relation to the group mobility, but each member sink can also move randomly within a certain group area in relation to the individual mobility. For supporting such groups, geocasting may decrease data delivery ratio due to continuous group area shifting by the group mobility, and multicasting may increase energy consumption due to frequent multicast tree reconstructions by the individual sink mobility. Recently, mobile geocasting protocols have been proposed, which enable a mobile sink group to periodically register its current group area information to a source and member sinks in the group to passively receive data from the source by flooding within the registered group area. However, due to the passive data dissemination, they suffer from excessive energy consumption of sensor nodes due to flooding data within the large group area and result in high data delivery failures of member sinks on edge of the group due to asynchrony between the registered group area and the actual group area. Therefore, we propose an active data dissemination protocol that exploits a local data area constructed by considering the moving direction and pattern of a mobile sink group. In the proposed protocol, a source sends data to nodes in the local data area in advance, and member sinks in the group actively receive the data from the local data area when they potentially pass it. To efficiently construct a local data area, we investigate the pattern of group mobility and classify into three major categories according to the prediction level: a regular movement, a directional movement, and a random movement. We then present three different data dissemination schemes with an efficient local data area to effectively operate for each mobility pattern. Experimental results conducted in various environments show that the proposed protocol has better performance than previous protocols in terms of the data delivery ratio and the energy consumption.