This research introduces LBTMA, a novel framework for effective traffic management in Internet of Things (IoT) networks employing software-defined networking (SDN). LBTMA comprises three modules: P4-enabled Stateful Traffic Monitoring (P4-STM), P4-enabled Distributed Load Balancing (P4-DLBS), and P4-enabled Distributed Packet Aggregation and Disaggregation (P4-DPADS). Operating entirely within the data plane, the three modules collaboratively address the challenges of managing high communication traffic from IoT devices. P4-STM utilizes state tables for flow monitoring and anonymization, while introducing a novel multi-controller communication scheme (MCCS) that separates routine data from critical alerts through two dedicated channels. MCCS demonstrated a 25% improvement in throughput and a 51% decrease in latency compared to single controller architecture. P4-DLBS features Enhanced Weighted Round Robin (P4-EWRR) load balancing algorithm, which leverages P4′s distributed decision-making capabilities and inter-switch coordination for enhanced scalability and reduced controller burden. P4-EWRR continuously adjusts server weights based on real-time factors (e.g., queue length, server resource pool, CPU utilization) to ensure efficient resource allocation. In testing, P4-EWRR achieved an average response time of 15 ms and an average packet drop rate of 2%. P4-DPADS employs a hierarchical data plane to efficiently handle high volumes of small IoT packets. It demonstrated an average disaggregation accuracy of 98%, communication overhead reduction rate of 70%, and an impressive average aggregation ratio of 95%. Additionally, P4-DPADS contributes to a 25% reduction in latency and a 40% increase in throughput. The LBTMA framework's modularity and P4 programmability provide flexible, scalable, and efficient traffic management in IoT networks.