Projects per year
Personal profile
Research interests
My research over the past 19 years has focused on an array of problems in Information and Data Sciences viewed through the lens of Information Theory, with an emphasis on both deriving fundamental performance limits and also on designing algorithms approaching these fundamental limits. A particular focus is on information storage/communication/processing systems that may be under attack by eavesdropping and/or jamming malicious parties – the tools I have helped develop over the last two decades provide unconditional informationtheoretic security guarantees (independent of cryptographic security guarantees often considered in security scenarios, which usually rely on computational hardness assumptions that are sometimes fragile).
My research group calls itself the CANDOIT team: Codes, Algorithms, Networks – Design and Optimization for Information Theory. The work that I and collaborators focus on lies in the intersection of and impacts, among other fields, Information Theory, Coding theory, algorithm design, highdimensional geometry, estimation theory, and optimization. Despite the tools of my trade being mathematically abstract and theoretical, they have tangible realworld implications and a broad range of datadriven applications, such as for largescale data processing, secure distributed computing, and robust distributed data storage.
Specific projects I am involved in, and which may potentially lead to research projects (feel free to email me), include:
Estimating sparse patterns from sparse data: Consider, first, the problem of grouptesting, which aims to identify the (hopefully) small number of individuals carrying a disease in a large population via as few "group" tests as possible (potentially due to lack of populationscale testing resources); these group tests involve pooling together samples from different subsets of individuals into a small number of pooled tests (each pooled test has a positive test outcome if and only if at least one individual whose sample was included was a disease carrier) and inferring the set of diseased individuals from the set of test outcomes. This problem of grouptesting is one example of nonlinear sparse signal estimation, wherein a sparse (mostly zero) input is to be inferred from a small number of outputs, where the inputoutput relationship is nonlinear. Fundamental limits (on the minimum number of tests required for reliable estimation) for the classical grouptesting problem have only recently been obtained; in this project we will investigate how to translate insights from these recent works to broad classes of estimation problems.
Fingerprint
 1 Similar Profiles
Projects
 3 Active
Research output

Communication Efficient Secret Sharing in the Presence of Malicious Adversary
Bitar, R. & Jaggi, S., 2020, In: arXiv preprint arXiv:2002.03374.Research output: Contribution to journal › Article (Academic Journal) › peerreview

Quadratically Constrained Twoway Adversarial Channels
Zhang, Y., Vatedka, S. & Jaggi, S., 2020, In: arXiv preprint arXiv:2001.02575.Research output: Contribution to journal › Article (Academic Journal) › peerreview

Covert Communication With Polynomial Computational Complexity
Zhang, Q., Bakshi, M. & Jaggi, S., 2019, In: IEEE Transactions on Information Theory. 66, 3, p. 13541384 31 p.Research output: Contribution to journal › Article (Academic Journal) › peerreview