The rapid deployment of IoT networks in different industrial services has caused the emanation of a huge volume of data from sensors and monitors. The efficient analysis and compact representation of the big data generated are critical to managing its transfer, storage, and latency. In this work, we propose a novel graphical model that has the flexibility to adapt to different architectural features of distributed IoT systems with low-power sensors. Further, we consider the traffic load minimization problem in the network and map the problems of network coverage and clustering onto the graph to find an optimal distributed network structure that minimizes the traffic load. The proposed approaches are tested under different network conditions, and the simulation results emphasize the improved performance of the method.