Hybrid localization using Wi-Fi and GPS signals (Machine learning)
Lin, Roy Weihao
Date of Issue2016-04-13
School of Computer Engineering
Centre for Computational Intelligence
In this report, the student first analyzes an existing Android App that does indoor navigation. An indoor localization machine-learning approach is then proposed by making use of the Received Signal Strength Indicators (RSSI) of nearby Wi-Fi access points. The algorithm has two phases. In the first phase, RSSI will be collected on an Android smartphone at Level 2, Section B, of the School of Computer Science and Engineering (SCSE) building at Nanyang Technological University. The data collected will be used as a training model for the second phase - testing. The testing phase would make use of three different machine learning algorithms (with kernel options) on the trained model. The algorithm with the least cumulative error on the estimated localized results would be most preferred.
Final Year Project (FYP)
Nanyang Technological University