![[PHOTO]](CN-2023.jpg)
|
CHANDRA NAIR
Professor
811, Ho Sing Hang (SHB)
| SOME QUICK LINKS |
My research interests and contributions focus on developing ideas, tools, and techniques to address families of combinatorial and non-convex optimization problems, primarily arising in the information sciences.
Since 2007, my work has been broadly motivated by the study of the optimality and sub-optimality of achievable regions in multi-user information theory—specifically, Marton’s achievable region for the two-receiver broadcast channel and the Han–Kobayashi achievable region for the two-sender, two-receiver interference channel.
Investigating Marton’s region has led to the development of a series of techniques, results, inequalities, and capacity regions by identifying extremizers of associated non-convex optimization problems. These methods are closely linked to concepts of sub-additivity and tensorization of functionals, prompting me to explore connections with hypercontractive inequalities and related topics in functional analysis.
While we have demonstrated the sub-optimality of the Han–Kobayashi region for the general two-sender, two-receiver interference channel, its optimality remains open for the important special case of the scalar Gaussian interference channel. This challenge has motivated us to investigate information inequalities with additive structures, leading to fruitful connections with inequalities in additive combinatorics.
In summary, my recent research interests revolve around deriving new families of information inequalities, particularly those exhibiting sub-additivity or inherent additive structures.
During my doctoral and post-doctoral studies, I focused predominantly on theoretical aspects of combinatorial optimization in both finite and large systems, often inspired by conjectures from statistical physics. In addition to these central themes, I have occasionally collaborated on a variety of related problems with several talented colleagues.
A summary of my research is available here.
TEACHING (recent)
Multiuser information theory (most recent: Spring 2025)
Signals and systems (most recent: Fall 2024)
Probability theory [graduate] (most recent: Spring 2024)
