MPAR: A Movement Pattern-Aware Optimal Routing for Social Delay Tolerant Networks

TL;DR

This paper introduces MPAR, a movement pattern-aware routing algorithm for SDTNs that optimizes relay node selection using search algorithms, significantly improving delivery performance over previous methods.

Contribution

The paper presents a novel MPAR routing algorithm that leverages movement patterns and search techniques to achieve global optimal relay node selection in SDTNs.

Findings

MPAR outperforms existing routing algorithms in simulations.

Tabu-MPAR effectively guides relay nodes to optimal sets.

Movement pattern analysis enhances routing efficiency.

Abstract

Social Delay Tolerant Networks (SDTNs) are a special kind of Delay Tolerant Network (DTN) that consists of a number of mobile devices with social characteristics. The current research achievements on routing algorithms tend to separately evaluate the available profit for each prospective relay node and cannot achieve the global optimal performance in an overall perspective. In this paper, we propose a Movement Pattern-Aware optimal Routing (MPAR) for SDTNs, by choosing the optimal relay node(s) set for each message, which eventually based on running a search algorithm on a hyper-cube solution space. Concretely, the movement pattern of a group of node(s) can be extracted from the movement records of nodes. Then the set of commonly visited locations for the relay node(s) set and the destination node is obtained, by which we can further evaluate the co-delivery probability of the relay node(s) set. Both local search scheme and tabu-search scheme are utilized in finding the optimal set, and the tabu-search based routing Tabu-MPAR is proved able to guide the relay node(s) set in evolving to the optimal one. We demonstrate how the MPAR algorithm significantly outperforms the previous ones through extensive simulations, based on the synthetic SDTN mobility model.