Visible to the public Biblio

Filters: Author is Kliewer, Jörg  [Clear All Filters]
2022-07-01
Günlü, Onur, Kliewer, Jörg, Schaefer, Rafael F., Sidorenko, Vladimir.  2021.  Doubly-Exponential Identification via Channels: Code Constructions and Bounds. 2021 IEEE International Symposium on Information Theory (ISIT). :1147—1152.
Consider the identification (ID) via channels problem, where a receiver wants to decide whether the transmitted identifier is its identifier, rather than decoding the identifier. This model allows to transmit identifiers whose size scales doubly-exponentially in the blocklength, unlike common transmission (or channel) codes whose size scales exponentially. It suffices to use binary constant-weight codes (CWCs) to achieve the ID capacity. By relating the parameters of a binary CWC to the minimum distance of a code and using higher-order correlation moments, two upper bounds on the binary CWC size are proposed. These bounds are shown to be upper bounds also on the identifier sizes for ID codes constructed by using binary CWCs. We propose two code constructions based on optical orthogonal codes, which are used in optical multiple access schemes, have constant-weight codewords, and satisfy cyclic cross-correlation and autocorrelation constraints. These constructions are modified and concatenated with outer Reed-Solomon codes to propose new binary CWCs optimal for ID. Improvements to the finite-parameter performance of both our and existing code constructions are shown by using outer codes with larger minimum distance vs. blocklength ratios. We also illustrate ID performance regimes for which our ID code constructions perform significantly better than existing constructions.
2020-06-02
Aliasgari, Malihe, Simeone, Osvaldo, Kliewer, Jörg.  2019.  Distributed and Private Coded Matrix Computation with Flexible Communication Load. 2019 IEEE International Symposium on Information Theory (ISIT). :1092—1096.

Tensor operations, such as matrix multiplication, are central to large-scale machine learning applications. These operations can be carried out on a distributed computing platform with a master server at the user side and multiple workers in the cloud operating in parallel. For distributed platforms, it has been recently shown that coding over the input data matrices can reduce the computational delay, yielding a tradeoff between recovery threshold and communication load. In this work, we impose an additional security constraint on the data matrices and assume that workers can collude to eavesdrop on the content of these data matrices. Specifically, we introduce a novel class of secure codes, referred to as secure generalized PolyDot codes, that generalizes previously published non-secure versions of these codes for matrix multiplication. These codes extend the state-of-the-art by allowing a flexible trade-off between recovery threshold and communication load for a fixed maximum number of colluding workers.