Resilient Service Embedding In IoT Networks (1910.05783v1)

Abstract: The Internet of Things (IoT) has been applied to a large number of heterogeneous devices and is used in the deployment of a variety of applications on the basis of its distributed open architecture. The majority of these IoT devices are battery-powered and are interconnected via a wireless network. IoT devices may be used to carry out critical tasks. Thus, the IoT network requires a resilient architecture that supports semantic search, failure discovery, data recovery, and dynamic and autonomous network maintenance. In this paper, we present a new resilience scheme for IoT networks. We evaluate the proposed scheme in terms of its power consumption and data delivery time, and then compare the results with those of recent resilience schemes such as schemes based on redundancy and replication. The proposed framework was optimized using mixed integer linear programming and real-time heuristics were developed, thus embedding a virtual layer into a physical layer based on a service-oriented architecture (SOA). The proposed framework offers different combinations of packet resilience in terms of recovering the lost data by using end- to-end mechanisms. We further analyzed these schemes by investigating the power consumption, data delivery time, and network overhead of these techniques. The results showed that the proposed splitting technique enhanced the network performance by reducing the power consumption and the data delivery time of service embedding by selecting energy-efficient nodes and routes in IoT networks.