For the management, operation, and optimization of smart networks, computational intelligence methods are desirable. In wireless sensor networks (WSNs), adding more nodes necessitates the seamless delivery of large amounts of data to different nodes. These high data transfer rates may cause wireless sensor networks to be overloaded, which provokes congestion, latency, and packet drop. In addition to causing information loss in wireless sensor networks, congestion also costs a lot. An effective computational intelligence strategy Mixed Routing Methodology (MRM) for improving data transfer while reducing latency is required to address this challenge is proposed. This schema able to route data packets by avoiding busy data path and selecting sensor nodes with maximum energy back up and lower data load. The proposed schema has two steps: Data route Formation and Data congestion-aware routing. Before to constructing the data routing, the residual energy of the sensor node is calculated. Then, the residual energy level is analyzed using regression manipulation to find whether or not the sensor node have sufficient energy backup.The energy-efficient nodal points are geographically positioned, and Multiple data paths between the source and sink sensor nodes will be Routed using route request and route reply. After that, congestion-aware routing based on buffer space and bandwidth capacity is performed using the MRM approach. Out of all possible routes, the one with the maximum bandwidth capacity and the least amount of buffer space is chosen as the best route. To minimize data latency and data loss along the data route path, congestion-aware data transmission is implemented. In relation to the quantity of data packets and sensor nodes, the simulation takes into account a number of performance parameters, including energy consumption, data delivery rate, data loss rate, throughput, and delay.

machne learning, heterogenous sensor network, routing.,