Fine-grained frequencies meet synchronous transmission

Abstract

Fine-grained frequencies have been used for IEEE 802.15.4 based communication in several recent works to enhance network throughput as well as combat Cross Technology Interference (CTI) in licence free ISM bands. However, the works demonstrate that as the Center Frequency Distance (CFD) goes very low (<3 MHz), adjacent channels start severely interfering with each other which heavily limits the use of the fine-grained frequencies. All the studies so far on fine-grained frequencies consider the underlying communication to be based on traditional Asynchronous-Transmission (AT). Synchronous-Transmission (ST) based communication, in contrast, has shown its superiority over AT-based strategies in several different aspects in many recent works. It has been shown that the ability to make efficient use of special physical layer phenomena called Capture-Effect (CE), enables ST-based strategies to carry out more in-parallel communication even under the same channel/frequency (under certain constraints), compared to AT. In this work, we study how this special capability of ST can be exploited in the context of multi-channel or multi-frequency communication with fine-grained frequencies, in particular, when the CFD goes below 3 MHz. We explicitly study how a very narrow band of frequencies can be effectively used for in-parallel communication. Based on extensive local and controlled experiment-based studies, we conclude that the construction of the groups plays a major role in exploiting the fine-grained frequencies. To fulfill the requirements, we design and develop a simple, distributed, and efficient group formation strategy, that can divide a given IoT-network into a desired number of groups and maximize the benefit of the conjunction of ST and the fine-grained frequencies. Through extensive experiment-based evaluations in IoT-testbeds, we show that the groups constructed by the proposed strategy can make very efficient use of the fine-grained frequencies with a CFD as low as 1 MHz to execute in-parallel communication tasks quite reliably and independently. We also demonstrate that even under severe CTI, when the traditional standard channel-based mechanism drastically fails or results in considerably poor performance, communication under fine-grained frequencies under the proposed mechanism can sustain pretty well.