The Internet of Things (IoT)-based model connects the innumerable devices which operate in a wireless mode to obtain the data information about the different attributes from their surroundings. The IoT devices suffer from the constrained energy sources; hence, it is utilized in an optimized scheme to extend the network lifetime as well as to enhance the different performance metrics. Hence, this research proposes the mean differential variation-based dung beetle optimization (MDV-DBO) approach for the energy-efficient, congestion-aware resource allocation and routing protocol (ECRAR) in IoT networks. The MDV-DBO approach effectively explores and exploits the solution space, making it scalable to larger IoT networks. Hence, the proposed MDV-DBO approach targeted to design an optimal routing strategy which considers the system configuration as well as traffic among the data in the networks. The effectiveness of the proposed MDV-DBO approach is validated through the various performance metrics. The proposed method attains the better throughput of 75%, energy efficiency of 81%, and network lifetime of 68% at the number of nodes of 200 respectively when compared to the existing methods like based multi-objective firefly optimization (MFFO).