Formal Development of Hierarchical Agent-Based Monitoring Systems for Dynamically Reconfigurable NoC Platforms

Formal Development of Hierarchical Agent-Based Monitoring Systems for Dynamically Reconfigurable NoC Platforms

Sergey Ostroumov (Åbo Akademi University, Finland) and Leonidas Tsiopoulos (Åbo Akademi University, Finland and University of Turku, Finland)
DOI: 10.4018/jertcs.2012040103
OnDemand PDF Download:
$37.50

Abstract

Sophisticated applications deployed on multi-core platforms require many resources as well as dynamic monitoring of the platform to provide efficiently and reliably the needed functionality. In this paper, the authors propose an approach to formal modelling with adequate tool support of an agent-based system whose function is to dynamically monitor the state of the multi-core platform and perform reconfiguration procedures under faults. For this purpose, the authors use the Event-B formalism which allows them to stepwise develop correct-by-construction specifications by mathematical proofs. Furthermore, the formalism enables the decomposition of a specification, which makes it possible to implement a well-structured and hierarchical agent-based monitoring system.
Article Preview

Agent-Based Monitoring System For Noc Platforms

The design of an agent-based monitoring and reconfiguration system requires special means to monitor the state of a system and take proper actions if inadequacies occur. Such means are usually implemented as agents (Rantala, Isoaho, & Tenhunen, 2007). When the number of functions performed by agents and/or the number of agents grows, it is reasonable to have a hierarchy of the agents, especially for a large-scale NoC platform. Generally, the hierarchy of the agents has a three-level structure as illustrated in Figure 1 (Yin, 2009).

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 8: 2 Issues (2017)
Volume 7: 2 Issues (2016)
Volume 6: 2 Issues (2015)
Volume 5: 4 Issues (2014)
Volume 4: 4 Issues (2013)
Volume 3: 4 Issues (2012)
Volume 2: 4 Issues (2011)
Volume 1: 4 Issues (2010)
View Complete Journal Contents Listing