In this thesis I research the ability of groups of agents to organize their collective behavior, without any human intervention. Using a framework for gathering information of the behavior, analyzing the performance, and updating the behavior, the agents can adapt to changing environments or user requirements. In my thesis I use different mechanisms driving the self-organization, but mostly focus on Distributed Constraint Optimization Problems (DCOPs) to do so. A new algorithm called CoCoA (Cooperative Constraint Approximation) is used to quickly find solutions that are near-optimal. Throughout my thesis the approach is put to use for different applications such as sensor networks, wireless power transfer networks and smart grids.@en

Wireless Power Transfer (WPT) technologies aim at getting rid of cables used by consumer devices for energy provision. As long distance WPT is becoming mature, the health impact of WPT becomes increasingly important to consider. In this paper we look at how to maximize the wireless power transfer to remote devices, while maintaining a safe level of electromagnetic radiation (EMR) for humans that are in the vicinity of the energy transmitters. Classically, this problem can be described as a centralized optimization problem of finding the optimal set of safe power levels at locations of human presence. Instead, we advocate to formulate this problem as an agent-based Distributed Constraint Optimization Problem (DCOP). As a solution to this problem we introduce CoCoA_WPT, a variant of the DCOP solver CoCoA. CoCoA_WPT provides a solution of similar quality to centralized solver even for a large scale network involving over a thousand nodes. Based on CoCoA_WPT, we propose a self-adaptive charging system: Transferring Energy Safely by Self-Adaptation (TESSA). TESSA keeps the charging network safe even when it is perturbed by environmental dynamics. We show that TESSA can reach on average up to 85% of the theoretical optimal maximum total transmitted power (calculated using centralized solution) while satisfying the EMR safety constraints.@en

We propose a novel incomplete cooperative algorithm for distributed constraint optimization problems (DCOPs) denoted as Cooperative Constraint Approximation (CoCoA). The key strategy of the algorithm is to use a semi-greedy approach in which knowledge is distributed amongst neighboring agents, and assigning a value only once instead of an iterative approach. Furthermore, CoCoA uses a unique-first approach to improve the solution quality. It is designed such that it can solve DCOPs as well as Asymmetric DCOPS, with only few messages being communicated between neighboring agents. Experimentally, through evaluating graph coloring problems, randomized (A)DCOPs, and a sensor network communication problem, we show that CoCoA is able to very quickly find solutions of high quality with a smaller communication overhead than state-of-the-art DCOP solvers such as DSA, MGM-2, ACLS, MCS-MGM and Max-Sum. In our asymmetric use case problem of a sensor network, we show that CoCoA not only finds the best solution, but also finds this solution faster than any other algorithm.@en