Sensor Network Management Through Fitness Function Design in Multi-Objective Optimization
Source: Syracuse University
Multi-objective optimization can support sensor network management by taking advantage of the many degrees of freedom available in controlling the sensors. Fitness function design is the key to increasing efficient use of the sensors complete a successful mission. This paper discusses applying fitness functions to model performance parameter decisions. Performance constraints can be introduced preventing solutions with fatal performance flaws from being considered as well as decreasing run-time. Also, the system's tolerance to missing performance goals may be increased or decreased by taking advantage of the weights in goal programming equations. The swarm can be designed to reduce run-time for real-time applications as well as improving the system's performance mismatches in key areas through the introduction of limits and performance weights in the fitness function.