<p>The Restricted Access Window (RAW) mechanism was introduced by the IEEE 802.11ah amendment to improve Dense Wireless Networks (DWN) performance, where a large number of nodes connect to a single access point. RAW groups nodes and schedules exclusive transmission slots for each group as a way of reducing competition for the wireless medium. Despite being a somewhat straightforward concept, some aspects of RAW can be improved to increase its performance in DWN, such as the optimization of the parameters—e.g., number of groups, length of the slots—and the definition of grouping criteria to avoid the hidden terminal problem. This paper presents a comprehensive survey on the state-of-the-art regarding RAW improvement strategies in DWN. In the literature, to the best of our knowledge, no other recent survey that addresses improvement strategies for the RAW mechanism. The surveyed papers span from parameter optimization to grouping algorithms and consider several different network performance objectives. In order to better organize the existing literature, we also present a taxonomy. The analyses showed that grouping algorithms improve network performance by reducing hidden terminals; however, some unresolved issues were identified, such as the lack of group balancing by these grouping algorithms. As a result, unbalanced groups intensify competition for medium access, leading to a higher number of collisions. The RAW parameter optimization proposed in the papers improved network performance both in terms of the evaluated metrics and QoS requirements, and the analytical models proved efficient in predicting network behavior. However, the evaluations were made through simulations, disregarding validation through real experiments.</p>