SDN-based fog and cloud interplay for stream processing

Adaptive Scheduling of Continuous Operators for IoT Edge Analytics

“…IoT data streams may be processed exclusively at the network Edge in [11,12], or in a hierarchical or a peer to peer (P2P) networks of Edge and Cloud nodes in [13,14,15,16,17,18,19,20]. EFC nodes are used in a P2P fashion in [8,21,22,23] or in a hybrid fashion combining P2P and hierarchical networks in [24,25,26]. Our work focus on a hierarchical architecture of EFC nodes that can be shared among several DSPA (or other) applications as also studied in [8,26].…”

“…EFC nodes are used in a P2P fashion in [8,21,22,23] or in a hybrid fashion combining P2P and hierarchical networks in [24,25,26]. Our work focus on a hierarchical architecture of EFC nodes that can be shared among several DSPA (or other) applications as also studied in [8,26].…”

“…Most of the works that rely on EFC (or Edge-Cloud) architecture focus on optimising the network resource usage and the DSPA response time by constraining only the computational resource usage [19,15,13,8,26]. In our work, a cost model is introduced to deal with the trade-offs of minimising the joint usage of both computational and network resources.…”

“…As shown in Challenges of Edge-Fog-Cloud paradigm. The EFC computing comes with the following challenges to overcome [8,9]:…”

“…• Simulation experiments: The proposed aTSOO-H algorithm is evaluated through extensive simulations based on real-world data sets. When compared with the baseline solution [10] and the solutions from related work [8], the results show that aTSOO-H achieves a high dynamic deployment success rate of DSPA applications, i.e., the best trade-off between resource usage cost, response time satisfaction and lower rescheduling cost. Rescheduling cost is defined as the difference between the number of operators that a DSPA application contains before after its scheduling.…”