Decentralized Channel Selection for Wireless Communication Systems Under Severe Out-of-Band Emission

This paper explores a channel selection problem for Diversified-Range IoT (DR-IoT), a wireless communication system that operates in a band with severe out-of-band emissions and has many channels with different bandwidths. Because DR-IoT operates in a band with severe out-of-band emissions, energy detection-based techniques do not work when probing to determine whether a network is operating in a given channel. Furthermore, hundreds of overlapping channels make channel scans difficult. To facilitate channel selection, DR-IoT employs an advertising channel (ACH) to exchange channel usage information among neighboring networks. If all networks can exchange their positions and current channels, they can assign channels to minimize the interference of signals sent from different networks. However, they often cannot exchange such information due to the limited ACH communication range. To mitigate this problem, we developed two new channel selection methods that leverage the ACH and received signal strength (RSS) from coordinators of other networks: minimum received beacon strength (MRBS) and closer neighbor further block (CNFB). CNFB is specifically designed to prevent networks from contending over selected channels, which coordinators cannot communicate with each other via the ACH. Our simulation results showed that communication on channels selected by CNFB achieves a 29.1 % increase in average throughput compared to a naive random selection method without ACH while outperforming the other reference methods.