<p>In the interconnected world of IoT, the ability to precisely pinpoint the location of nodes is a fundamental cornerstone. Without accurate node localization (NL) in Sensor Networks, the data collected by sensors often lacks context and irrelevance because of the absence of information about the sensor nodes' locations. Node localization in IoT states the method of defining the physical position of IoT devices or nodes within a network or environment. The accuracy of sensor node positioning plays a major role in determining the overall performance of Sensor Networks. Received Signal Strength Indicator (RSSI) is indeed a profitable approach for assessing the position of a non-anchor node in sensor node localization systems. In this article, the Coati Optimization Algorithm (COA) and its modified forms are practiced to enhance the localization accuracy of nodes localized. To improve the accuracy and randomness of the coati optimization algorithm, two approaches, chaotic map and levy flight are applied on COA and named these algorithms as C-COA and LF-COA. The proposed fitness function not only focused on the current value of distance but also take care of the past average distance. The result evaluation and performance analysis of COA, C-COA and&#xa0;LF-COA shows that LF-COA performs better that the rest of two algorithms, COA, C-COA.</p>