WSN lifetime optimization through controlled sink mobility and packet buffering

Maximizing the lifetime of energy constrained wireless sensor networks is a very challenging issue. It has been demonstrated that the use of a mobile sink can significantly increase the network lifetime by balancing the load among nodes. However, in existing solutions, nodes either send their data through multihop towards the sink which induces energy consumption due to relaying process, or the nodes store the data until the sink comes at their vicinity, which usually requires an infinite buffer capacity or induces buffer overflow if the buffer is of fixed size. In this paper, we propose a new approach in which nodes send their data through multihop path of reduced length by offering nodes the possibility to buffer data while waiting the sink coming closer (not necessarily at node range), which exempts more sensors from relaying these data. This strategy allows to save energy, while ensuring no data is lost due to buffer overflow. We model the problem of optimizing WSN lifetime with limited buffer capacity and controlled mobile sink using a Linear Program (LP). For arbitrary topologies, our LP determines the sink sojourn times at each possible location, the data transfer rates between nodes and the buffered packets quantities. Compared to previous models, our solution achieves better lifetime and enables to generate and transmit more data to the mobile sink. We show that our scheme also better balance the load among nodes. Finally, we derive from the numerical results a distributed algorithm for data collection in WSN. We show through simulation that our approach leverages latency for energy saving, and we compare its performances against single-hop and multihop forwarding protocols.