The proliferation of mobile devices in our society has brought huge amounts of data traffic over the radio airwaves. A fundamental problem to ensure that data can be smoothly transported by the wireless media is effective control and mitigate interference. This project advanced the field of interference control and management through the following dimensions.  We gained fundamental understanding of the underlying physics of some new interference management (IM) algorithms, including the tradeoffs between performance and overhead. We developed tractable mathematical abstraction to grasp their physical behaviors. A key feature of our models is that they are implementable at the physical layer. We developed some new IM algorithms that removed the limitations of the state-of-the-art.  We studied various optimal deployment scenarios of IM algorithms for wireless Internet of Things (IoT).  Some representative research outcomes from this project are:

(i)  We developed new theoretical results on how MIMO degrees of freedom (DoFs) at a node are used for spatial multiplexing (SM) and interference cancellation (IC) under general channel rank conditions. Our theoretical development focused on a single SM link and IC on a single interference link and then extended to multiple links.  We found that a shared DoF consumption for IC at both transmit and receive nodes is most efficient for DoF allocation under rank-deficient conditions. This result fills in a critical gap in the existing DoF models under full-rank conditions.

(ii)  We further developed new DoF IC models that exploits interference signal strengths in the eigenspace. Instead of performing IC with DoFs on all directions in the eigenspace as in existing DoF models, we showed the benefits of performing IC with DoFs only on those directions with strong signals in the eigenspace. To differentiate strong and weak interference within an interference link, we introduced the concept of effective rank threshold. IC is only performed for strong interference corresponding to large singular values in the eigenspace based on this effective rank threshold while weak interference is treated as noise in throughput calculation.

(iii)  We developed a new technique to reduce the computational burden of singular-value decomposition (SVD) in hybrid beamforming (HB) by exploiting sparsity at mmWave channels. Specifically, we use only a small number of the most significant singular vectors of a matrix, derived from randomized SVD technique. By limiting operations only to the key information of our interest and judiciously choosing a proper target rank for lower rank approximation, we can drastically reduce the complexity involved in the traditional beamforming.

(iv)  We developed a novel 5G scheduler-mCore---that can achieve ~1 ms scheduling with joint optimization of RB allocation and MCS assignment to MU-MIMO users. The key idea of mCore is to perform a multi-phase optimization, leveraging large-scale parallel computation. In each phase, mCore either decomposes the optimization problem into a number of independent sub-problems, or reduces the search space into a smaller but most promising subspace, or both. Experimental results showed that mCore can offer better performance when compared to other state-of-the-art algorithms.  

The project offered education and training opportunities to a group of graduate students. Due to the technical depth of this project, the students were required to develop cross-disciplinary expertise by taking courses or receive online training in electrical engineering (wireless communications), operations research (optimization theory), and computer science (algorithm design and parallel computing). Such expertise from diverse areas is very unique to the graduate students who participated in this project. Some students in the project have produced multiple high quality articles that have been either published or accepted by major journals and conferences.  Throughout this project, the students worked closely with the PI as well as a team.  Several students had the opportunity to work as summer interns in industry, which helped them position their research in a practical setting.  Student who participated in this project were highly valued by industry and have found rewarding career opportunities in industry upon their graduation. 

Last Modified: 01/17/2022
Modified by: Thomas Hou