IBM Research Zurich · Historical manuscript

Bandwidth Adaptive Multicast for Pervasive Monitoring

A 2005-era research design for changing the detail and bandwidth of monitoring updates according to application and recipient needs.

Authors: Aditya Mohan, François Dolivo, Dirk Husemann, Reto Hermann.
Institution: IBM Research Zurich and Brown University, circa 2003–2005.
Type: Research manuscript.

The engineering problem

Distributed monitoring systems often receive state information from many devices, each with different network capacity and information needs. Sending every update at full precision to every recipient wastes bandwidth, while aggressive reduction risks hiding changes that matter.

The proposed approach

The coauthored manuscript explores bandwidth-adaptive multicast for monitoring: represent system state at different levels of resolution, propagate changes using group communication, and tailor update fidelity to resource constraints. The goal is to conserve bandwidth while keeping useful monitoring information available.

The manuscript presents a proposed approach to bandwidth-adaptive multicast for pervasive monitoring.

Connection to the wider research program

The study complements Communication through Almost Anything, which investigates physical transport paths when conventional radio is obstructed. The two manuscripts address different layers of a monitoring system: physical connectivity and information dissemination.

More generally, both belong to the historical 1999–2005 record, not to present-day Robometrics® AGI results.

Related research

View the bibliographic record →