ZigBee is a protocol for high-level communication under IEEE 802.15.4 standard. ZigBee network specification creates a personal area network for small, less-power consuming radio communication. The vast use in device to device communication of ZigBee turns its possibility in the use of the Internet of Things. A large amount of sensor nodes is interconnected through the Coordinator in ZigBee network. The end nodes send a huge amount of sensing data to the Coordinator via intermediary nodes like routers. These huge data creates overhead application traffic load in routers and Coordinator. A load balancing mechanism is necessary for a ZigBee network to contain multiple Personal Area Networks (PAN) and switch the threshold PAN loads with simplified network routing. This paper represents a novel ZigBee network model based on Stochastic and Widest Path process for proper traffic orientation, queuing and analyze load minimization issues. The two statistical models are put through in the ZigBee network using Riverbed to reduce and balance the overhead load. The result indicates that applying Stochastic process minimize the Medium Access Control (MAC) layer load as well as the overall ZigBee network load. This research will be helpful in large-scale integration of sensor nodes in a balanced ZigBee network for data load.