5 Conclusions and future work
State-of-the-art data suppression methods aim at obtaining a predetermined accuracy level of the collected information for an entire period, when a monitoring task is executed. In target tracking applications, the required accuracy of collected target position information varies significantly in time. A reduced accuracy is often sufficient for making optimal sink decisions. The proposed suppression approach exploits the above insight to decrease the amount of data transmitted in WSN. Instead of suppressing transmissions of sensor readings that are not necessary to accurately estimate target position, the introduced method suppresses those sensor readings that are not useful for selecting optimal movement direction by the mobile sink.
According to the proposed approach, at each time step of the tracking procedure the mobile sink has to minimize its distance to a location in which the target can be caught. The data describing actual target position are useless and should be suppressed, if there is a high probability that the above mentioned distance will be minimized when the sink will move towards a recently reported target position instead of the actual target position.
Results of the simulation experiments clearly show advantages of the proposed approach. When comparing with stateof-the-art algorithms, the decision-aware suppression allows the data communication costs (packet rate and hop count) to be significantly reduced without decreasing performance of target tracking. Furthermore, the decision-aware suppression enables a beneficial trade-off between the tracking performance (time to catch) and the data communication cost. The negative effect of communication failures on the tracking performance can be effectively mitigated for this approach by using the ARQ method.
