Enhancing Quality of Service in Cluster-Based IoT Routing Using Optimized Lion Routing with Deep Deterministic Policy Gradient (OLRDDPG)

The combination of limited energy resources, changeable network topology, and extensive heterogeneous deployments, the quick expansion of Internet of Things (IoT) networks has created major issues in guaranteeing Quality of Service (QoS). Reducing latency, optimizing data delivery, and extending network lifetime all depend on effective routing. Conventional routing protocols and independent metaheuristic optimization methods frequently have poor performance in dynamic environments and little adaptability. This research suggests Optimized Lion Routing with Deep Deterministic Policy Gradient (OLRDDPG), a unique hybrid routing architecture for cluster-based IoT networks, as a solution to these problems. To accomplish adaptive and QoS-aware routing, the suggested method combines the continuous decision-making power of Deep Deterministic Policy Gradient (DDPG) with the global exploration capabilities of the Lion Optimization Algorithm (LOA). In the suggested model, DDPG dynamically optimizes routing choices based on network conditions including residual energy, delay, and connection quality, whereas LOA is used for effective route discovery and population initialization. Key QoS parameters, including as End-to-End Delay, Packet Delivery Ratio (PDR), Routing Overhead, Throughput, and Energy Consumption, are used to assess OLRDDPG's routing performance. The suggested OLRDDPG greatly outperforms current methods like Glowworm Swarm Optimization (GSO), Shuffled Frog Leaping Algorithm (SFLA), Convolutional Lion Routing Optimization (CLRO), and Deep Belief Lion Optimization (DBLO), according to extensive simulation findings. The findings demonstrate that OLRDDPG is a good fit for large-scale, dynamic IoT applications because it reduces latency and energy consumption while increasing network throughput and dependability.