Self-Optimised Tree Overlays Using Proximity-Driven Self-Organised Agents

Abstract

Hierarchical structures are often deployed in large scale distributed systems to structure communication. Building and maintaining such structures in dynamic environments is challenging. Self-organisation is the approach taken in this chapter. AETOS, the Adaptive Epidemic Tree Overlay Service, provides tree overlays on demand. AETOS uses three local agents to this purpose (i) to translate application requirements to self-organisation requirements, (ii) to self-organise nodes into optimised tree topologies based on these requirements, (iii) to control bootstrapping and termination of self-organisation. The evaluation of AETOS in different simulation settings shows that it provides high connectivity in tree overlays optimised according to application requirements.

Decentralized Planning of Energy Demand for the Management of Robustness and Discomfort

“…MIN-INTERVENTIONS does not have a significant influence on robustness. Note that as the number of possible plans increases, robustness also increases in average 52%, for l = f 1 (z = 2), 61% for l = f 1 (z = 3), 30%, for l = f 2 (z = 1), and 39%, for l = f 2 (z = 2), confirming earlier findings concerning equivalent possible plans [16], [7].…”

“…AETOS self-organizes agents in different random positions for each tree topology to capture the effect of topological positioning. The effect of different types of tree topologies is evaluated in earlier work [16], [7].…”