Robust and Tuneable Family of Gossiping Algorithms

TL;DR

This paper introduces a family of gossiping algorithms with a tunable parameter that controls communication parallelism, allowing adaptation to system conditions for improved performance and collision avoidance.

Contribution

It proposes a flexible, parameterized gossiping algorithm family and methods to dynamically tune the communication level based on system resources.

Findings

Algorithms can be tuned for optimal communication parallelism.

Dynamic parameter adjustment improves robustness against message collisions.

Enhanced performance in varying communication environments.

Abstract

We present a family of gossiping algorithms whose members share the same structure though they vary their performance in function of a combinatorial parameter. We show that such parameter may be considered as a "knob" controlling the amount of communication parallelism characterizing the algorithms. After this we introduce procedures to operate the knob and choose parameters matching the amount of communication channels currently provided by the available communication system(s). In so doing we provide a robust mechanism to tune the production of requests for communication after the current operational conditions of the consumers of such requests. This can be used to achieve high performance and programmatic avoidance of undesirable events such as message collisions.