EE4C03 STATISTICAL DIGITAL SIGNAL
An indoor, line of sight and robust localization algorithm using a microphone array in square arrangement was developed to localize and track a toy car in an indoor environment. Using an audio beacon on the car and channel estimation and peak finding algorithms, the TDOA can be obtained from the input of the microphones. With this TDOA, a location estimate is made. Incorrect measurements or outliers are discarded using a static algorithm, and a path prediction algorithm maps the actual route of the target.
The results show that due to modeling issues and very noisy data, the channel estimation algorithm is unable to correctly estimate the channel, resulting in location estimates which are completely wrong. However, using simulated TDOA values, location estimation and path prediction are shown to be effective.
It is concluded that the channel estimation and peak finding algorithm work according to theoretical research and as feedback, an improved measurement environment is recommended.