IEEE 802.11 PSM in ns-3: A Novel Implementation and Evaluation under Channel Congestion

With the proliferation of high-throughput Wi-Fi devices, low-power Wi-Fi enabled devices, such as Internet of Things (IoT) sensors, that generate sparse or sporadic traffic must co-exist in increasingly dense wireless network conditions. A key technology that enables Wi-Fi as a connectivity solution for low power operation is the Power Save Mechanism (PSM). While the Wi-Fi module in Network Simulator-3 (ns-3) is detailed, it lacks an implementation of PSM to simulate and study the behavior of devices utilizing PSM. In this paper, we develop an open-source model for simulating PSM in ns-3 that supports uplink (UL) and downlink (DL) unicast and multicast traffic and maintains proper PHY layer states. We couple our implementation with a power model that accounts for energy consumed during state transitions and use this simulation framework to evaluate the performance of PSM with sparse UL traffic under high channel congestion with increasing number of nodes competing for channel access. We show that with 20 contending stations with high load, the average current and awake time of a Power Saving (PS) station with sparse UL TCP (Transmission Control Protocol) traffic increase by as much as 108% and 254% respectively. We also show that the bidirectional nature of TCP traffic increases the average current and awake time of the PS station by 25% and 75% respectively with 10 contending stations.