Application of Quantum-Clustering on Thermograms of WiFi Circuits in Different Operation Modes

摘要

The purpose of this work is to evaluate the efficacy of applying a model-based quantum clustering (QC) algorithm on thermograms of functional modes in WiFi circuits. As unsupervised clustering algorithm, it can work on clusters of any shape and does not require any prior information. QC proves its efficacy for many applications, it has been tested, in this work, and compared with other algorithms which suffer randomness according to initialization. The tests are conducted on thermograms of an electronic chip in different operation modes. The benefits of QC are confirmed through performance analysis of clustering algorithms. Robustness analysis is also conducted against white-Gaussian noise clustering and so on classification of actual WiFi circuit operation modes based on thermograms.