Wireless sensor network (WSN) is deployed in number of applications for real-time data collection. Sensors that are smaller, cheaper, and more sophisticated have made this possible in recent years. These sensors are outfitted with wireless interfaces through which they can connect with one another to build a network. In the WSN, routing plays an important role to determine the optimal path to successfully deliver the data from source to destination node. In the literature, two types of routing available such as static and dynamic. Out of these, dynamic routing is preferred in the current scenario due to robust, low cost, and node position is dynamic. Further, dynamic nodes are classified into three protocols such as proactive, reactive, and hybrid protocols. Therefore, in this paper, initially, these routing protocols are examined. After that, we have studied and analysed the various research papers are published in these routing protocols. Finally, open research challenges are defined based on the study.

Due to APTEEN’s practicality, the routing protocol has become a popular choice. However, there are issues with the network’s unequal energy consumption, the early death of certain nodes, and the channel’s poor effective coverage. In the literature, biologically motivated algorithms have been implemented for optimization of the versatile and edge-sensitive energy potential networks routing protocol. These algorithms are iterative in nature. As a result, in general, choosing the best cluster head takes a lengthy time to execute. To solve these problems, this paper uses Huffman coding and artificial neural networks to optimize the Adaptive Threshold-sensitive Energy Efficient Network (APTEEN) routing protocol. The Huffman coding reduces the data packet size, whereas an artificial neural network searches for the optimal cluster head in the APTEEN routing protocol. The proposed method is simulated in MATLAB and performance evaluation is done based on alive nodes, residual energy, First Node Dead (FND), Head Node Dead (HND), and Last Node Dead (LND). The result shows that the proposed method has higher alive nodes, residual energy, and better FND, HND, and LND nodes than the existing algorithms such as EDS-APTEEN and GA-APTEEN.