An Efficient Message Collecting and Dissemination Approach for Mobile Crowd Sensing and Computing

Abstract

Mobile Crowd Sensing and Computing (MCSC) is an emerging technology along with the popularity of mobile devices. We utilize the concept of Delay Tolerant Networks (DTNs) and edge/fog computing to support the message collection and dissemination for the MCSC. The biggest challenge here is to design an efficient routing method to deliver messages for both “upload” (data collection to edge nodes) and “download” (data dissemination to nodes that interest) paths. We assume that the mobile crowd nodes will like to exchange data in a DTN manner based on opportunistic transmission in order to save energy and data transmission cost. We design a probability-based algorithm to upload data carried by normal mobile nodes to the edge nodes. Then, we use cosine similarity to relay specific message of attributes to users who have high interest to receive the message. We simulate the algorithm with the National Chengchi University (NCCU) real trace data of campus students, and compare it with other traditional DTN routing algorithms. The performance evaluations show the improvement of message delivery ratio and decreasing latency and transmission overhead.