Adaptive transmit power control based on signal strength and frame loss measurements for WLANs

In the past few years, we witnessed a rapid penetration of Wireless Local Area Networks (WLANs) into the home and enterprise. Emerging technology such as the IEEE 802.11n radio, which is getting increasingly affordable, makes delivering multimedia content over wireless networks possible and this would drive the technology further into our daily life. As the number of available wireless channels in the unlicensed spectrum is limited (3 non-overlapping channels in 2.4GHz unlicensed band and up to 24 non-overlapping channels in 5GHz unlicensed band), they have to be shared by multiple WLANs consisting of Access Points (APs) and STAtions (STAs). In a Multi-Dwelling Unit (MDU) WLAN deployment, e.g. in an apartment building or hotel, transmissions in overlapping cells tend to interfere with each other. This will adversely impact the aggregate wireless network throughput and thus the quality of experience for applications such as multimedia streaming. Hence there is a need for automatic and adaptive resource management strategy to ensure a good overall network performance.;In this thesis we propose an adaptive per-link Transmit Power Control (TPC) solution for WLANs. TPC can reduce interference, increase channel reuse, and eventually increase the overall capacity in dense 802.11 wireless networks. However intelligent algorithms are required to adapt transmit power in a practical and distributed way to achieve improvement in performance. It becomes more challenging given different types of interference (cooperative and non-cooperative) in the unlicensed band as well as the hidden node problem. From a detailed study of the previous efforts at power control, we observe that in order to make better decisions on transmit power; an AP needs to actively monitor several factors. Hence we develop a TPC algorithm based on both link margin estimation as well as frame loss rate measurement. Compared to previous solutions that adapt the transmit power based on measurement of a single parameter (either received signal strength or frame loss rate), the proposed power control mechanism can diagnose and take remedial action for hidden nodes and channel access asymmetry problems manifesting as frame losses. It is adaptive to mobility, complementary to any rate control algorithm and can also be incrementally deployed amidst non-cooperative nodes. We have implemented the algorithm as an application running on Atheros chipset-based 802.11n APs, taking practical system-level limitations into account. The proposed solution achieves significant transmit power reduction at the APs (to as low as 60% of the maximum power) for STAs as far as 70ft and over ∼60% increase in total network throughput through interference mitigation.