Wireless Sensor Networks (WSNs) present various hurdles, particularly in terms of energy efficiency, due to the sensor nodes' limited power resources. By utilizing interactions across the many layers of the communication protocol stack, cross-layer optimization has become a viable method for addressing these issues. An extensive review of cross-layer optimization techniques created especially to improve energy efficiency in WSNs is provided in this study. We classify these protocols according to the layers with which they interface and the optimization goals they aim to achieve, such as throughput maximization, latency reduction, and energy consumption minimization. We also examine the fundamental design ideas and methods used in these protocols, including cross-layer feedback mechanisms, adaptive modulation and coding schemes, combined routing, and MAC layer optimization, and more. By means of an organized analysis and contrast. We evaluate and compare current cross-layer optimization procedures in a methodical manner to determine their advantages, disadvantages, and possible directions for further investigation. In addition to highlighting the need of cross-layer techniques for attaining energy-efficient operation in WSNs, our analysis offers guidance for the creation of more reliable and scalable solutions in this field. In this paper we proposed a classifier for saving energy efficiently named as EEDCA (Ensembled Enhanced Distributed Channel Access which is a combination of ML+WSN to solve some basic issues like energy optimization and also increase the performance. The evaluation's findings demonstrate that using our suggested strategy results in a system that performs better and conserves more energy.