Best-effort delivery of emergency messages in post-disaster scenario with content-based filtering and Priority-enhanced PRoPHET over DTN

Abstract

Despite concerted efforts for relaying crucial situational information, disaster relief volunteers experience significant communication challenges owing to failures of critical infrastructure and longstanding power outages in disaster affected areas. Researchers have proposed the use of smartphones, working in delay tolerant (DTN) mode, for setting up a peer-to-peer network enabling post disaster communication. However, due to opportunistic contacts, intermittent connectivity and constrained resources (mainly battery) of mobile devices, not all messages get forwarded in such a network. As a result, several crucial messages, waiting in the buffer may get dropped and unimportant, extraneous (sentimental comments of victims, etc.) messages may get transmitted at the cost of expensive network resources. Therefore, filtering out such irrelevant messages (containing unrelated information and emotional expressions) and prioritizing relevant messages according to its importance becomes crucial so that messages which are critical for decision making receive the highest priority and get disseminated in the network at the minimum possible time. In this paper, our objective is to ensure best-effort delivery of emergency messages using a two-step approach, (i) segregation of high priority messages through natural language processing based filtering and (ii) dissemination of filtered messages over DTN using a Priority-enhanced PRoPHET routing protocol which is developed by adapting the popular PRoPHET routing protocol. The protocol successfully achieves on-the-fly message prioritization and ensures best effort delivery of such prioritized messages to their appropriate destinations. The proposed protocol is evaluated using the ONE simulator. Simulation results suggest that our protocol outperforms other well known DTN routing protocols like PRoPHET, Spray-And-Wait, MaxProp and Epidemic in terms of delivery ratio of prioritized messages and overhead ratio.