The large range of applications for autonomous drone navigation, including search and rescue, aerial surveillance, and package delivery, has made it a prominent research topic in recent years. In comparison to flying drones through challenging terrain, navigating robots is very simple. DeepReinforcementLearning (DRL) has received a lot of interest since it has the experience learning ability to perform challenging tasks with little knowledge. Navigation is a drone's main fundamental difficulty. Deep reinforcement learning (DRL), which enables the drone to learn from its experiences and develop its decision-making abilities, has emerged as a viable technology for autonomous drone navigation. The literature on DRL-based algorithms, architectures, and applications for drone navigation is reviewed in this article. Additionally, it identifies the gaps in the existing body of research and analyses the difficulties with DRL-based techniques. Future directions for this field of study are discussed as the review comes to a close. Overall, this systematic review indicates potential for additional study in this area and offers a thorough grasp of the current state-of-the-art in DRL-based techniques for autonomous drone navigation