The Characteristics of Multi-Criteria Weight on Ad-Hoc Network with Ant Colony Optimization

There are many types of weighting methods widely applied in solving multi-criteria problems in various fields of life. This paper analyzes three weighting methods namely rank-sum (RS) weights, rank order centroids (ROC) weights, and equal weights. The three weights are used to find out their performance in the selection of path pairs on multi-criteria ad hoc networks assisted by ant colony optimization (ACO). The multi-criteria are signal to noise ratio (SNR), load variance, and power consumption. The analysis shows that ROC weights have the best performance compared to RS weights and the equal weights based on load variance, power consumption, and path pair criteria. For the SNR criteria, the RS weights and the equal weights have better performance than the ROC weights at the SNR value in the range of 43.1 dB - 44 dB. However the SNR value in the range of 44 dB - 45 dB then equal weights have the best performance compared to RS weights and ROC weights.