Wireless Sensor Network emerges as a solution to replace the large network deployment. WSN grow due to the low power embedded systems and substitute the huge network infrastructure. However, the Data collection and data mining play a vital role in decision-making during congestion. In the existing WSN systems, congestion controls obtain through intra-clustering and Omnidirectional antenna. In modern congestion control, WSN applies with multi cross-layer techniques to avoid infrastructure cost and to increase the efficiency of packet delivery. In the proposed system, we apply probability control algorithm with three-tier logic to control congestion. The three tier logic works in between the MAC and Network layer for efficient congestion control through varying the flow control of data size and rate. The flow control of data size and rate maximize the throughput and packet delivery with low power consumption.