|Kaidong Wang||University of Southern California, USA|
|Konstantinos Psounis||University of Southern California, USA|
802.11ax introduces OFDMA to WiFi. It thus enables multiplexing users/user groups in the frequency domain. WiFi networks usually operate in a multipath environment which generates a frequency selective channel. Hence, the capacity of a user/user group changes over different subcarriers. A good scheduling and resource allocation scheme can maximize the sum rate by allocating users and user groups on subcarriers based on their CSI and other system considerations. In this paper we investigate how to optimally assign users and user groups to subcarriers with the goal of maximizing the user sum rate in the context of 802.11ax. We introduce a novel divide and conquer based algorithm which we prove to be optimal under the assumption that a user can be assigned to more than one resource unit (RU) which consists of one ore more subcarriers. This serves as a tight upper bound on the actual problem where users/user groups can be assigned to a single RU only per the 802.11ax standard. We then introduce two practical algorithms for the actual problem, a greedy one and a recursive one which jointly splits the bandwidth into RUs and schedules users on them. Extensive simulations comparing the performance of the aforementioned algorithms establish that our practical schemes achieve very good performance in all studied scenarios.