By spite of extensively researched and applied in many fields, direction of arrival (DOA) estimation remained many challenging problems in military applications, especially on the Internet of Battlefield Things (IoBT) systems. Therefore, this study presented a survey on the direction of arrival estimation methods applied for the IoBT. Differing from the industrial Internet of Things (IoT) systems, the IoBT systems were usually jammed or leaked information by the enemy. For eliminating the mentioned problems, an optimal adaptive antenna array could help to manage the transmitting space and energy. Hence, it can prevent the signal interception and jamming from the enemy electronic warfare systems. An overview of IoBT, including its limitations in practice, was expressed to understand the extraordinary challenges that it needs to resolve. Then, signal models for both general and particular arrays were formulated to understand the mechanism of DOA estimation methods. The selected well-known DOA estimation methods for popular antenna array configurations were mathematically explained, simulated, and compared. The simulation and comparison results indicated that each approach method has its advantages and disadvantages in use. In particular, the conventional beamforming was stable for DOA estimation, but it has lower accuracy and resolution caused by wide beamwidth. The MVDR method improved the DOA estimation accuracy and resolution, but it required an invertible covariance matrix. Finally, the MUSIC method and its extended ones achieved the best results in term of accuracy and resolution of DOA estimation. However, the MUSIC method required a knowledge of the number of incoming signal sources to estimate DOA. All techniques meet a problem of un-distinguishability in case of coherent signals, whereas the number of coherent sources is higher than three.