Energy-aware clustering of ubiquitous devices for collaborative mobile applications

The continuous advances in mobile computing technologies have paved the way to the spreading of new classes of collaborative mobile applications. Clustering the mobile nodes into local groups may promote and ease collaborations. In this paper we propose a clustering scheme formulated as a multi-criteria optimization problem based on the weighted-sum approach. The aim of the optimization is to maximize the network lifetime and the mobility similarity of the devices within a group. Depending on the specific application scenario, the weights associated to the cost functions can be properly tuned. Experimental results show that by using the proposed approach, the network lifetime is extended and the number of alive devices is significantly higher compared to alternative clustering strategies, while meeting application-level performance constraints as high mobility similarity within a cluster.