Pramanick, Subhendu2025-11-122025-11-122025-1052p.http://hdl.handle.net/10263/7623Dissertation under the supervision of Dr. Pratyay Mukherjee & Dr. Debrup ChakrabortyAbstract Through the distribution of secret key information among several parties, threshold cryptography improves the security of cryptographic systems by preventing any one entity from possessing the entire secret key and requiring a threshold number of participants to carry out cryptographic operations. This paradigm not only mitigates single points of failure but also ensures fault tolerance in the presence of compromised or unavailable parties. The Distributed Symmetric-key Encryption (DiSE) framework, introduced by Agrawal et al., realizes Threshold Symmetric-key Encryption (TSE) by requiring interactive participation from a threshold subset of servers for each encryption or decryption operation. While DiSE and similar TSE schemes provide strong security guarantees, they become inefficient for large datasets, as each encryption or decryption requires a separate round of server interaction, leading to significant computational and communication overhead. To address these scalability challenges, Christodorescu et al. proposed Amortized Threshold Symmetric-key Encryption (ATSE), which allows a privileged client to encrypt a large collection of messages with a single interaction with the key servers. Importantly, decryption still requires threshold server participation for each ciphertext, ensuring that all clients receive the same level of privacy and authenticity as in DiSE. The ATSE framework is built upon the novel primitive of Flexible Threshold Key-derivation (FTKD), enabling the interactive derivation of pseudorandom keys in a threshold-secure manner. This dissertation extends the ATSE paradigm to tabular data, presenting a new scheme that enables efficient and secure encryption of entire tables. Unlike traditional TSE schemes, which incur high overhead by requiring server interaction for each data cell, our construction allows a privileged client to encrypt an entire table with a single round of interaction, while maintaining strong security guarantees for individual decryptors. This advancement significantly improves the practicality and scalability of threshold encryption for modern, data-intensive applications.enThreshold cryptographyThreshold Symmetric-Key Encryption for Tabular DataThesis