ABSTRACT
The Internet of Things (IoT) covers the way for creating connected infrastructures pervasively to support better efficiency and flexibility with innovative services. Recently, such advantages are becoming attractive for the industrial domain also, where the IoT systems (known as Industrial IoT (IIoT)) are entering into the industry marketplace with proposed designed solutions. However, due to the rapid growth of mobile Internet services and wireless communications, it is required to enhance the performance of the present IoT technology to achieve low power consumption, low cost, enhancement of coverage area, and so forth. In this direction, narrowband IoT (NB-IoT), which was started by Third Generation Partnership Project (3GPP), has acquired popularity as it can be deployed in 5G long-term evolution (LTE) networks directly on a low-frequency channel bandwidth of 180 kHz and low data rate (∼ 250 kbps). Due to such low channel bandwidth and low data rate, NB-IoT is unsuitable for IoT-based industrial applications related to the delivery of prioritized traffic with delay sensitivity; and therefore, packet scheduling is a crucial aspect of quality of service (QoS) requirements in NB-IoT–based IIoT networks. By considering this challenge, in this paper, we design online learning–based adaptive packet scheduling of prioritized traffic, which aims to reduce packet transmission delay in the NB-IoT framework. The performance analysis of the proposed narrowband-prioritized traffic scheduling (NBPTS) through simulation shows that it can significantly improve the network performance in the NB-IoT paradigm.
