QoS-Aware Resource Allocation for Mobile IoT Pub/Sub Systems

Abstract

Abstract. IoT applications are usually characterized by large-scale demand and the widespread use of mobile devices. Similarly, performing interaction among application and system components in a decoupled and elastic way, and enforcing Quality of Service (QoS) usually also become issues. Hence, paradigms such as pub/sub on top of cloud resources represent a suitable strategy for application development. However, management of QoS-aware resource allocation for pub/sub systems remains challenging, especially when system peers connect in an intermittent way. In this paper, we propose a new approach for resource allocation focusing on end-to-end performance in face of peers' disconnections. We evaluate and demonstrate the benefits of our approach using simulations. QoS enforcement was achieved in almost all scenarios, and we have shown that our approach can help reasoning about efficient resource allocation.

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

“…In [6], we provide a QoS-aware solution for allocating resources to pub/sub brokers by taking into account the end-to-end response time between peers, as well as their intermittent connectivity. As shown in Fig.…”

“…It is worth noting that in [6] we ignore possible application layer event processing, which is targeted in the approach proposed here.…”

“…That is, multiplexing is more complex at the middleware level, and we need to investigate how to address this issue in the system model. Similarly, regarding the treatment of intermittent connectivity proposed in [6], and as suggested in Section 2, we need to investigate whether a simple extension of the use of ON/OFF queues to the application level (to model the voluntary disconnections of mobile services and IoT devices) would be consistent with the treatment of disconnections at the middleware level, especially on an end-to-end basis.…”

“…This is in contrast with reactive approaches, such as in [5] and [4], which start with a initial (default) resource configuration and adapt it as QoS violations are detected. Thirdly, as introduced in previous work [6], resource allocation at the middleware level takes into account the possibility of intermittent connectivity of services, thus contributing to make the approach feasible for service choreographies that involve mobile services and IoT devices. Finally, resource allocation takes into account the fact that services can be shared across multiple choreographies, enabling the optimized allocation of resources [7].…”

Performance Analysis of Internet of Things Interactions via Simulation-Based Queueing Models

“…To model QoS settings such as the intermittent connectivity [20] of mobile things, a separated queueing model was investigated with the aid of an ON/OFF queueing model [16,17]. The applicability of our models was demonstrated in [16][17][18][21][22][23] for the performance of publish/subscribe and data-streaming systems, respectively. In [19], we applied our approach to analyze the performance of IoT devices with heterogeneous QoS settings.…”