Moving centroid based routing protocol for incompletely predictable cyber devices in cyber-physical-social distributed systems

Cyber-Physical-Social Distributed Systems (CPS-DS) represent the combination of cyber world, physical world and social world. Cyber devices, such as mobile sensors, with characteristics of self-organization and lack of infrastructure, have great influence on data transmission and aggregation in CPS-DS. Incompletely Predictable Networks (IPNs) are a specific representation of CPS-DS, where cyber devices move in limited ranges around basic positions. In this paper, we proposed a novel protocol named Moving Centroid based Routing Protocol (MCRP). Note here that moving centroid is an adjustable point constructed by a series of positions, which are centroids of a set of discrete particles, recording experienced positions of a node during past time slots. Moving centroid is utilized to improve the performance of the proposed protocol, considering movement trend of individuals in a network. A node generates an adjusted basic position (ABP) message, which contains the centroid of a set of locations composed with node positions at every known time slot. ABP messages are periodically broadcast to update the network topology of basic positions. After obtaining the up-to-date topology of basic positions, data packet sender selects optimal relay node through transmission probability calculated by basic positions. From the simulation results, the proposed protocol performs well with ABP messages being broadcast every 20 s. In particular, compared with AODV and GrD-OTBR, the performance of our protocol has a promotion in terms of packet delivery ratio and delay in networks with high node density.