Distributed scheduling using belief propagation for internet-of-things (IoT) networks

The number of internet-of-thing (IoT) devices has recently been growing at a rapid rate. From the fact that most of IoT devices are connected through advanced wireless technologies, their coexistence issues become important. Distributed and densely-deployed nature of IoT networks render wireless scheduling very challenging. This work develops a distributed scheduling strategy for a network of wireless IoT devices. To be precise, it aims at maximizing the overall sum rate of the wireless network where a centralized coordination is not supported. The proposed approach considers a synchronized slotted structure consisting of two phases: distributed scheduling and distributed communication phase. In the distributed scheduling phase, IoT devices, via reciprocal exchange of simple messages, share local information with neighboring devices and decide scheduling policies. In the distributed communication phase, the devices communicate with their neighbors on scheduled slots. To this end, a state-of-the-art message-passing framework is introduced to develop a distributed scheduling algorithm. Based on the notion of a factor graph, the developed distributed scheduling algorithm finds an efficient scheduling solution that maximizes the overall sum rate of the network. Simulation results verify that the developed algorithm outperforms existing distributed techniques to a considerable extent in a consistent fashion.