DTA

Digital Theses Archive

 

Tesi etd-09262017-195613

Thesis type
Dottorato
Author
CALVARESI, DAVIDE
URN
etd-09262017-195613
Title
Real-Time Multi-Agent Systems: challenges, model, and performance analysis
Scientific disciplinary sector
INF/01
Corso di studi
INGEGNERIA - Ph.D. Programme in Emerging Digital Technologies (EDT)
Commissione
relatore BUTTAZZO, GIORGIO CARLO
Presidente Prof. DI NATALE, MARCO
Membro Prof. DRAGONI, ALDO FRANCO
Membro Prof. ARSLAN, TUGHRUL
Parole chiave
  • real-time multi-agent systems
Data inizio appello
;
Disponibilità
completa
Riassunto analitico
In the last decades, the Multi-Agent Systems (MAS) resulted in being one of the most relevant approaches fostering the development of systems performing distributed thinking and reasoning, automated/autonomous actions, and regulating component interactions in unpredictable and uncertain scenarios. <br>Some practical examples are intelligent and pervasive systems (assistive monitoring and feedback in telerehabilitation), energy management, and energy negotiation.<br>Those application domains particularly involve three major characteristics: intelligence, autonomy and real-time behavior. <br><br>Although MAS are one of the major technological paradigms used to implement such systems, they mainly address intelligence and autonomy but miss to comply with strict timing constraints.<br>The compliance with strict-timing constraints (real-time compliance) is crucial for safety-critical applications operating for example in the healthcare, industry 4.0, and automotive domains. The main reasons for this lack of real-time satisfiability in MAS originate from current theories, standards, and technological implementations. In particular, traditional internal agent schedulers, communication middlewares, and negotiation protocols have been identified as co-factors inhibiting the real-time compliance. This manuscript provides an analysis of such MAS components and paves the road for achieving MAS capable to meet strict-timing constraints, thus fostering reliability and predictability.<br><br>Moreover, by studying possible adoptions of MAS in Cyber-Physical Systems (CPS), it has been observed that in current MAS, the actual task execution is still delegated to traditional general-purpose scheduling algorithms running within the agent (local scheduler of behaviors). The main consequence is the incapability to enforce compliance with strict timing constraints (impossibility of providing any guarantee about the system’s behavior in the worst-case scenario). Therefore, the adoption of MAS is hampered, excluding significant application scenarios such as “safety-critical environments&#34;. <br>Another contribution presented in this manuscript is the schedulability analysis of various tasks-sets, feasible using real-time schedulers, on top of traditional general-purpose solutions. In particular, the study of deadline-missing rate occurring in general-purpose setups is evaluated on a self-developed agent-based simulator (named MAXIM-GPRT) developed on OMNET++. The results strengthen the motivations for adopting and adapting real-time scheduling mechanisms as the local scheduler within agents.
File