Edward Knightly, professor and chair of electrical and computer engineering, Ph.D. student Oscar Bejarano, and collaborators Eugenio Magistretti, recent Rice PhD graduate, and Omer Gurewitz, former Rice postdoctoral student and professor at Ben Gurion University, won Best Paper at the 2014 SECON conference in Singapore.
“The work is on Multi-user MIMO in IEEE 802.11ac-based networks and the efficiency of the sounding procedure. Hopefully, this work will influence the way channel feedback acquisition is performed in the standard,” lead author Oscar Bejarano said.
Multi-user MIMO is a technique that leverages multiple antennas at a transmitter to serve multiple devices simultaneously. For example, an Access Point (AP) that serves multiple users at the same time instead of sequentially, one at a time, as it is done in legacy systems.
In general, an access point consists of a device that allows other wireless devices (users) to connect to the wired infrastructure, and commonly, to the Internet. Examples of user devices are cell phones, laptops, other compatible equipment.
“With MU-MIMO, the AP can serve as many users as it has antennas,” Bejarano explained.
With MU-MIMO, the capacity of the network can potentially be increased proportionally to the number of concurrently served users. MU-MIMO requires the transmitter to have information about the channel between all its transmit antennas and all the users, and it obtains the information via a process called “sounding”.
Sounding is typically done before every single MU-MIMO transmission and it is very time consuming, which means that the actual overhead to enable MU-MIMO is very high. The team proposed a solution in which the AP can alleviate this issue.
“MUTE is a protocol we have implemented to make MU-MIMO more efficient. It strives to minimize, or temporarily eliminate the excessive overhead associated with sounding. With MUTE, we try to avoid sounding whenever it is not critical. To this end, we allow the AP to learn about the channel statistics of all users in order to make a prediction about how much their channels have changed. Consequently, if the system recognizes that certain channels have remained relatively stable, the AP adjusts accordingly, therefore reducing the amount of time spent sounding the channel,” Bejarano said.
Their paper showed that in their testing they were able to use MUTE to reduce the overhead sounding creates by almost 73%.
“The goal with MUTE is to make MU-MIMO much more efficient and push the real system-level gains close to the gains shown in theory,” Bejarano added.
Congratulate the team here.