Passive Sonar systems make use of the target noise received from the acoustic sensor array to determine the target distance and angular. The linear three-element array (TEA) technique is one of the most popular approaches among the sensor array ranging systems, where the three sensors of array are arranging on an equal-distant line. Recently, to improve the flexibility of target ranging system, the three sensors are suggested to arrange on a line and distance between sensors with none equidistant, and is referred to as the non-linear TEA ranging technique. For obtaining accurate information of angular of the Target sources using the TEA ranging, it is highly dependent of the result of time-delay estimation (TDE). Consequently, the selection of the TDE techniques become very significant, when both of linear and non-linear TEA ranging approach is employed. In this paper, among various adaptive TDE techniques, we focus on the one with the adaptive hybrid LMS (HLMS) filtering algorithm in the DCT domain, for extracting the information of time delay between related sensors. In consequence, we could apply it for the Target sources ranging associated with the TEA approaches. The merits of the above-mentioned approach are verified with computer simulations. We show that it outperforms the one with other conventional approaches for obtaining the information of angular target sources, which can be employed for target ranging for the next step.