Cognitive Networks: Towards Self-Aware Networks

1.3. Swarm Intelligence

In biology, the intelligent and well-organized behavior of a group of social insects can often be observed. For example, ants or bees solve complex tasks like building nests or searching for food by distributing simple tasks among each other. In such emergent systems, it is not so much the individual work which determines the outcome, but rather the collection of all single activities. This behavior is generally referred to as swarm intelligence ^[]. The collaboration of insect societies is based on the principle of division of labor where specialized workers perform specialized tasks in parallel, leading to a better efficiency. Division of labor can be seen for all insect types in the distinction of worker and reproductive castes. In some cases, several different types of workers coexist which have distinctive tasks as well, e.g., protection of the nest or nurturing the brood.

Each individual insect makes its own decisions, solely based on locally available information. In other words, the insect is not aware of the global situation but only gains its input by interaction with other members of its species. This interaction may be either by direct contact (visual or chemical) or indirect interaction. In the latter case, one individual influences the environment and another reacts to this change at a later time. Ants, for example, interact with each other by laying a trail of pheromones which influences other ants to follow this trail, thus, reinforcing it. ...