Source author record

Ivo Kubjas

Ivo Kubjas appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Information Theory math.IT Distributed, Parallel, and Cluster Computing

Catalog footprint

What is connected

2works

3topics

1close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2020arXiv

Failure Probability Analysis for Partial Extraction from Invertible Bloom Filters

Invertible Bloom Filter (IBF) is a data structure, which employs a small set of hash functions. An IBF allows for an efficient insertion and, with high probability, for an efficient extraction of the data. However, the success probability of the extraction depends on the storage overhead of an IBF and the amount of the data stored. In an application, such as set reconciliation, where there is a need to extract data stored in the IBF, the extraction might succeed only partially, by recovering only part of the stored data. In this work, the probability of success for a partial extraction of data from an IBF is analyzed. It is shown that partial extraction could be useful in applications, such as set reconciliation. In particular, it allows for set reconciliation by using the IBF, where the storage overhead is too small to allow full extraction. An upper bound on the number of rounds in an iterative set reconciliation protocol is presented. The numerical results are derived analytically, and confirmed by the computer simulations.

preprint2015arXiv

Data Dissemination Problem in Wireless Networks

In this work, we formulate and study a data dissemination problem, which can be viewed as a generalization of the index coding problem and of the data exchange problem to networks with an arbitrary topology. We define $r$-solvable networks, in which data dissemination can be achieved in $r > 0$ communications rounds. We show that the optimum number of transmissions for any one-round communications scheme is given by the minimum rank of a certain constrained family of matrices. For a special case of this problem, called bipartite data dissemination problem, we present lower and upper graph-theoretic bounds on the optimum number of transmissions. For general $r$-solvable networks, we derive an upper bound on the minimum number of transmissions in any scheme with $\geq r$ rounds. We experimentally compare the obtained upper bound to a simple lower bound.