GRU based optimal restricted access window mechanism for enhancing the performance of IEEE 802.11ah based IoT networks.

IEEE 802.11ah, known as Wi-Fi HaLow standard, specifically promoted for next-generation Internet of things (IoT) applications. Restricted access window (RAW) mechanism is introduced in IEEE 802.11ah medium access control (MAC) layer. The RAW mechanism reduces the collisions among contending devices by partitioning RAW period into RAW sots and allocates a RAW slot to each group. Thus, the specific group of devices alone contend for channel access in the respective RAW slot. In this paper, we develop an accurate analytical model to compute the throughput and energy efficiency of IEEE 802.11ah with the RAW mechanism. In IEEE 802.11ah, the choice of optimal number of RAW slots is an open research problem. Thus, we present a gated recurrent unit (GRU) based deep learning-recurrent neural network (DL-RNN) to estimate the optimal RAW slots that improves the performance of IEEE 802.11ah for dense IoT networks. From the results, we observe that throughput and energy efficiency performance are significantly improved by using optimal number of RAW slots obtained using GRU. The analytical works conducted are validated with extensive simulation works. [ABSTRACT FROM AUTHOR]