Predicting impact of Hitchhike on coexisted heterogeneous IoT networks

Abstract

In Internet of Things (IoT), plenty of heterogeneous nodes coexist and contend for wireless channel for data transmission. Therefore, improving the channel utilization is of great importance. Hitchhike is a promising mechanism of improving the channel utilization. With Hitchhike, a node can carry control signals on the preamble field of another transmitting packet, thereby eliminating the extra time of transmitting the control signals and improving the channel utilization. However, this control signals may seriously interfere with the packet transmission of coexisted heterogeneous networks. In this paper, considering that an 802.15.4 network with Hitchhike enabled and an 802.11 network coexist, we develop theoretical models to predict the impact of Hitchhike on the time and frequency synchronizations, the packet error rate, as well as the system throughput of the 802.11 network. Extensive experiments verify that the proposed prediction models are very accurate and effective. Experimental results show that when control signals are superposed on the 802.11 Preamble, they can hurt the 802.11 performance (specifically, frame synchronization, coarse frequency estimation, and system throughput). However, they have little effect on symbol timing synchronization. This study is very helpful for providing parameter optimization and protocol-design guidance for the coexistence of heterogeneous IoT networks. When multiple heterogeneous networks coexist, we may apply machine learning algorithms to help design better Hitchhike protocols. The insights from our theoretical model (say, the frequency spectrum of the Hitchhike control message is a good design parameter for minimizing the interference with coexisted heterogeneous networks) can be used to choose appropriate machine learning algorithms and construct appropriate algorithm modules, for improving the performance of Hitchhike protocols in complex IoT scenarios.