CPS: Cross-interface network Partitioning and Scheduling towards QoS-aware data flow delivery in multimedia IoT
Hua Qin,Weimin Chen,Ni Li,Tao Wang,Hao Chen,Gelan Yang,Yang Peng
DOI: https://doi.org/10.1016/j.jnca.2023.103698
IF: 7.574
2023-08-01
Journal of Network and Computer Applications
Abstract:In the Internet of Things (IoT), multimedia traffic for audio, image, and video accounts for the largest proportion (over 78.7%) of the total traffic, bringing forward the vision of multimedia IoT (M-IoT). As part of the realization of loT, M-IoT is a general network paradigm that constitutes many smart objects equipped with the capability to collect multimedia data from the physical environment and deliver the data to other things. To satisfy a certain level of user experience, Quality of Service (QoS) is required to be regulated to ensure acceptable delivery of the multimedia content. As the most widely-used wireless technology, WiFi has been recommended for IoT communications for its high data rate, native IP compatibility, and good reusability of the existing infrastructures. However, WiFi suffers from channel contention, especially during multi-hop communications, which degrades the QoS performance and hinders its use for many M-IoT services. Although numerous protocols have been proposed to mitigate WiFi contention, they often consume much WiFi bandwidth for network control, lowering the level of achievable QoS performance. To address this issue, we propose a distributed Cross-interface network Partitioning and Scheduling (CPS) protocol, which leverages the co-existing ZigBee communications to divide the network into partitions and allows only one node in each partition to use its WiFi interface to transmit data at any time, for bandwidth-efficient and delay-constrained data flow delivery in M-IoT. A prototype node is implemented by integrating COTS ZigBee and WiFi interfaces into a BeagleBone Green wireless platform for IoT. Extensive field experiments are conducted in a multi-hop network of 24 prototype nodes that deliver real multimedia data (images and videos). The experiment results show that CPS outperforms the standard WiFi and a state-of-the-art contention control scheme (by 62.6% and 26.4% under high data traffic, respectively) in terms of a QoS metric capturing two basic performance metrics (i.e., bandwidth efficiency and end-to-end delay) of multi-hop communications, while retaining fair QoS performance and high energy efficiency.
computer science, interdisciplinary applications, software engineering, hardware & architecture