DRANS: Daily Routine Analysis for Node Searching in delay tolerant networks

Xingyu He,Guisong Yang,Hao Zhang

doi:10.1177/1550147717718105

Abstract

In delay tolerant networks, the success rate and the transmission speed are restricted by limited social interaction and complex node mobility pattern analysis. To increase the success rate and reduce the transmission delay in delay tolerant networks, we propose Daily Routine Analysis for Node Searching in delay tolerant networks. In Daily Routine Analysis for Node Searching, each node is required to generate a Staying Probability Table and a Transiting Probability Table by analyzing its own daily routine, then to distribute its Staying Probability Table and Transiting Probability Table to the whole network with the help of other nodes having different mobility patterns. On the basis of the Staying Probability Table and Transiting Probability Table, Daily Routine Analysis for Node Searching further provides a node tracking strategy and an opportunistic routing strategy for delivering data from the source node to the destination node. Trace-driven experiments are performed to compare Daily Routine Analysis for Node Searching with previous node searching methods. The experimental results demonstrate that Daily Routine Analysis for Node Searching is able to promote the success rate and reduce the transmission delay effectively.

Highlights

With the popularization of mobile devices in all kinds of application scenarios, delay tolerant networks (DTNs) as a novel type of communication paradigm have attracted lots of interests from the academic to the industry.[1]
Daily Routine Analysis for Node Searching (DRANS) with three forwarding paths and three forwarding nodes is denoted by DRANS1, while DRANS with five forwarding paths and five forwarding nodes is denoted by DRANS2
We compare DRANS with DSearch[15] and TSearch[16] (TSearch without and with encounter recording exchanges are denoted by TS1 and TS2, respectively) in the following four aspects: (1) success rate: the percentages that the source nodes successfully deliver data to their destination nodes, (2) average delay: the average time cost by the source nodes to deliver data to the destination nodes, (3) average transmission overhead: the average number of packets transmitted among nodes, and (4) average memory usage: the average amount of memory unit used by each node

Summary

Introduction

With the popularization of mobile devices in all kinds of application scenarios, delay tolerant networks (DTNs) as a novel type of communication paradigm have attracted lots of interests from the academic to the industry.[1]. In TSearch,[16] nodes can play different roles, have preferred subareas, and share encounter information and meeting probabilities, so all the three factors, such as meeting probability, social relationship, and geographical information, are taken into consideration in routing decision Even these node searching methods do bring some effective and positive influence on success rate promotion and transmission delay reduction; there are still some problems need to be solved: (1) in DTNFLOW,[14] the transiting probability is obtained by predicting and the node flows between subareas are not time related, which may not be reliable for real-time node searching; (2) in DSearch,[15] it is not practical and convenient to require each node to leave a hint about where it will go before it leaves its current subarea, because a node may change its route at any time or not surely know where it will go; and (3) in TSearch,[16] it may happen that all friends and encounters of a source node know nothing about its destination node, which means the help from friends and encounters is limited and cannot make sure every node is always traceable for other nodes. The performances of DRANS are evaluated and analyzed in section ‘‘Performance evaluation.’’ In section ‘‘Conclusion,’’ we conclude the work and give a future research direction

Related work

Experimental setup

Experimental results

Conclusion