Cross-layer QoS-Aware Resource Allocation for IoT-Enabled Service Choreographies

Fábio M. Costa (1) , Nikolaos Georgantas (2) , Raphael Gomes (3) , Ricardo da Rocha (4) , Georgios Bouloukakis (2, 5)

1 UFG - Universidade Federal de Goiás [Goiânia]
2 MIMOVE - Middleware on the Move
3 IFG - Instituto Federal de Goiás [Câmpus Goiânia]
4 Instituto de Biotecnologia Universidade Federal de Catalão
5 Donald Bren School of Information and Computer Sciences

Resource allocation for distributed systems is a well-known approach to deal with quality of service requirements. However, existing approaches do not consider the effects of resource allocation at the different levels of a system, especially when considering the end-to-end behavior of distributed compositions such as service choreographies. Based on our previous research on the optimized and QoS-aware resource allocation for service choreographies and publish-subscribe brokers, we outline a novel approach that integrates the treatment of resource allocation at both levels in order to yield a more precise control over end-to-end QoS. The paper presents an early architecture and discusses some of the main challenges towards realizing the approach.

Mots clés: Middleware; Service choreographies; QoS; Resource allocation

Domaines: Génie logiciel [cs.SE]; Calcul parallèle, distribué et partagé [cs.DC]; Modélisation et simulation; Systèmes et contrôle [cs.SY]

Fábio M. Costa, Nikolaos Georgantas, Raphael Gomes, Ricardo da Rocha, Georgios Bouloukakis. Cross-layer QoS-Aware Resource Allocation for IoT-Enabled Service Choreographies. The 5th International Workshop on Middleware and Applications for the Internet of Things (M4IoT), Dec 2018, Rennes, France. ⟨hal-01943541⟩