9.4. Sensory and Actuator Functions

The MAS characteristic of 'situated' encompasses the ability of agents to sense their environment as well as take actions on their environment. The ability to sense the environment provides the reasoning method with up-to-date information on the state of the network. These sensory inputs also provide the feedback necessary for learning. Actuator functions provide the ability to enact the solution or decision reached after reasoning. Therefore, sensory and actuator functions are vital to learning and reasoning.

Sensory functions lead to parallels between cognitive networks and sensor networks. Sensor networks are deployed to observe changes in an environment. In a cognitive network, the environment consists of the network itself. The cognitive network can be conceptually thought of as having a sensor network embedded within it that interacts with the cognitive elements through the SAN API (see Figure 9.1) to provide the status of the network. Sensor networks are usually considered to be limited in terms of available energy and processing power at each sensor node. The energy constraint drives sensor networks to limit communication in the network. In the cognitive network, it is also desirable to limit communication in the network, though the reason is to allow the communication channels in the network to serve users and applications as much as possible.

The processing limitations are less applicable to the sensors within cognitive networks. However, ...

Get Cognitive Networks: Towards Self-Aware Networks now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.