Title: Toward Scalable Software-defined Massive MIMO Wireless Networks
Abstract: Abstract: Massive MIMO is considered a core technology for 5G and beyond, promising large throughput gains and a significant increase in power efficiency and communication range through beamforming. There is a significant push to realize Massive MIMO using off-the-shelf computing platforms to speed up deployment and reduce costs. However, building a software-defined Massive MIMO network, while desirable, remains a significant challenge due to its enormous system hardware and computational complexity.
In this talk, I will first discuss our effort in the design and real-world deployment of a software-defined massive MIMO wireless network. I will discuss how our effort led to the rollout of the first public massive MIMO testbed that is available to the research community. Second, I will then discuss our work in the development of a Massive MIMO baseband processing real-time software. I discuss how the rich data parallelism in massive MIMO baseband processing can be utilized to scale the processing to many-core CPU servers and to achieve real-time performance and low latency. I will also talk about the use of this software on the public testbed and how it can impact massive MIMO research in the community. I conclude by providing an outlook on the future of software-defined massive MIMO wireless networks and the potentials and challenges ahead.
Speaker Bio: Rahman Doost-Mohammady is an assistant research professor in the Department of Electrical and Computer Engineering at Rice University.
His work focuses on the design and efficient implementation of next-generation wireless network systems. He is the technical lead for the Rice RENEW project. He received his Ph.D. from Northeastern University in 2015, M.S. from the Delft University of Technology in 2009, and B.S from Sharif University of Technology, in 2007, all in Computer Engineering.