Improving MIMO Performance in Wi-Fi Networks by using Collision Resolution and User Selection

Abstract

Multiple-Input Multiple-Output (MIMO) technologies have emerged as a key component to increase the capacity of wireless networks. The MIMO scheme either simultaneously transmits to multiple users at a time or focuses energy towards a single user to enhance the data rate. A number of Wi-Fi standards based on MIMO technology have been developed, and recently, several commercial products have been successfully deployed on the market. Unfortunately, many commercial MIMO-based Wi-Fi products fail to fully exploit the advantages of the MIMO technology, even though the MIMO technology could play a key role in improving the wireless network performance. MIMO nodes cannot provide their higher data rates, especially when they coexist with SISO nodes. Meanwhile, in Wi-Fi networks, significant Channel State Information (CSI) feedback overhead has been obstacle to the performance of MU-MIMO transmission and user selection. Most of these problems are observed to root in the inefficient PHY and MAC design of current MIMO based Wi-Fi systems: the MAC simply abstracts the advancement of PHY technologies as a change of data rate. Hence, the benefit of new PHY technologies are either not fully exploited, or they even may harm the performance of existing network protocols. In this dissertation we introduce three co-designs of PHY/MAC layers for MIMO based Wi-Fi networks, in order to overcome the intrinsic limitations of the current MIMO based Wi-Fi network and improve the network capacity. First, we show the Interference Alignment and Cancelation (IAC) based collision resolution scheme for heterogeneous MIMO based Wi-Fi systems. Second, we present a practical user selection scheme for MU-MIMO Wi-Fi networks. Finally, we improve the proposed user selection scheme by exploiting a frequency domain signaling scheme and using a capacity gain as a selection metric. We have validated the feasibility and performance of our designs using extensive analysis, simulation and USRP testbed implementation.