LAUSR

In the design of routing protocols for opportunistic networks (OppNets), the node’s context information such as number of encounters to destination (EN), distance to destination (DI), delivery probability (PR), to name a few, is used as routing decision patterns when selecting the best next hop candidate to forward the data packets to their destination. Most routing protocols thus far proposed for OppNets have considered either of these patterns or a combination of few of them as design objectives in their next hop selection processes. But none of these work have ever addressed their optimization for the same. In this regard, this paper proposes a novel multi-objectives based technique for optimized routing (MOTOR) in OppNets. This technique involves the use of a weighted function to decide on the next hop selection of a node based on a combination of objectives, namely, maximizing the number of encounters (EN), maximizing the delivery probability (PR), and minimizing the distance to destination (DI). A non-dominated set of solutions is proposed using a Naive and Slow algorithm for forwarding the data packets towards their destination. Simulation results are provided to assess the performance of the MOTOR scheme when the next hop selection process relies on a single objective (EN, PR, or DI), double objectives (EN–DI, EN–PR, or DI–PR), and triple objectives (EN–DI–PR). It is shown that the performance of the proposed routing scheme under the triple objectives option is better than that obtained under all the three double objectives in terms of delivery probability, average latency and number of messages dropped. For instance there is 6%, 41.84% and $$29.26\%$$29.26% increase in delivery probability, average latency and number of messages dropped respectively in triple objectives with respect to the double objectives DI–PR.