WIRELESS SYSTEM GENETIC ALGORITHM For successful communication, the radio must meet three conditions: • The radio must be configured to operate in the correct channel condition • The radio must support services such as voice and user-specified data• The radio must meet all regulatory requirementsTo successfully meet all three conditions, the radio must be aware of its environment and stimuli. It needs a cognitive engine that can analyze the physical link (channel), user requirements and regulatory requirements. Therefore, the cognitive engine must balance multiple objectives and constraints. To develop such a cognitive engine, a multiple-objective genetic algorithm is used. This is known as Wireless System Genetic Algorithm (WGSA). In this algorithm, the radio behavior is defined as a set of PHY and MAC level parameters and modeled as genes of a chromosome. Also general parameters such as payload size, voice coding, equalization, diffusion technique etc. they are modeled as genes. This model is then optimized using evolutionary processes as defined in the previous section. The WGSA defines chromosome fitness with a set of weighted fitness functions. Each fitness function is weighted based on its relative importance to the user. Therefore, the optimal solution is the one that provides the maximum QoS effectiveness for the user's requirements given the input constraints defined above. The suitability evaluation functions represent the current link quality of PHY and MAC layers which includes average transmit power, data rate, bandwidth, interference suppression, packet latency, packet jitter, BER, packet error speed (PER) and spectral efficiency. An important aspect of the WGSA is that each fitness function can eit...... middle of paper ......les/publications/Fortuna_Mohorcic.pdfA cognitive network is a communication network augmented by a knowledge plane. This KP extends vertically across OSI layers and horizontally across network technologies and nodes. The KP contains these two main elements: • Knowledge of the network scope, i.e. devices, architecture, technology, etc. • A cognitive cycle as defined in the previous section. Therefore, we can define a cognitive network as a network with a cognitive process contained in its Knowledge Plane. This process allows the network to sense current network conditions, act on those conditions, learn from the consequences of these actions, and at the same time achieve end-to-end user-defined high-level goals. The Knowledge Plane also builds a global perspective of the network and shares this information as needed with nodes in the network.