*** Data underlying the publication "Using global optimization methods for three-dimensional localization and quantification of incoherent acoustic sources" ***
Authors: Bieke von den Hoff
Promotors: prof.dr.ir. M. Snellen, prof.dr. D.G. Simons

Aircraft Noise and Climate Effects Group, Faculty of Aerospace Engineering, Delft
University of Technology

Contact Information:
m.snellen@tudelft.nl
Delft University of Technology - Faculty of Aerospace Engineering

***General Introduction***
This dataset underlies the paper "Using global optimization methods for three-dimensional localization and quantification of incoherent acoustic sources".
The paper aims to identify sound sources in terms of location and source level through a global optimization method.

The data shared here is the outcome of the global optimization code. The code was written for in-house use by prof. dr. ir. Mirjam Snellen and therefore can be requested with the prof.

***Purpose of the test campaign***
The campaign specified several different configurations of three sources in an anechoic room to create a controlled multi-source multidimensional system.
The true source locations are enlisted in the paper. The comparison with conventional beamforming was performed by co-author Roberto Merino-Martinez.

***Test equipment***
CAE Bionic M112 array which includes 112 MEMS microphones and an optical camera at the centre.

***Description of the data in this data set***
1. The data set contains the '.mat'-files of the individually measured sources which is the baseline level used in the figures.
2. 'postprocessing_GOResults.m': The processing code is supplied, for which the path of the dataset has to be included prior to running the code.
3. Zip-archives are provided with the GO results for measurements with the CAE array, different sources, and different cases.
   These archives contain a '.mat'-file with the output of the global optimization algorithm, in which the naming specifies the settings of the algorithm.
4. 'CSM_z=diff.mat' contains the measured Cross Spectral Matrix (fully calibrated) that is necessary for acoustic imaging
5. 'CSM_z=diff.mat' contains the measured Cross Spectral Matrix (fully calibrated) that is necessary for acoustic imaging
6. 'Mic_coordinates.mat' contains the coordinates of the microphones that are necessary for acoustic imaging and the microphones that are removed from processing ('bad_mics')

An example to explain the naming convention of the data files in the zip-archives:

results_calibrated_CAE_S1S2S3_z=diff_l=same_incoh_q=64_F=06_pc=075_Nsrc=3_Nruns=5_Ngen=600_23Hz

which indicates the following:
- Results of the specific case
- Calibrated data according to a specified frequency domain calibration
- Measured with the CAE array
- Sources that were active during recording (in example above: source 1, source 2, source 3)
- 'z' indicates whether the sources had the 'same' or a 'diff/different' normal distance to the array
- 'l' indicates whether the sources had the 'same' or a 'diff/different' level 
- incoh/coh indicates if the sources were initialised with incoherent or coherent signals
- q, F, pc, Nsrc, Nruns, Ngen indicate the setting parameters of the algorithm used to create the specific result (see also t he publication for explanation)
- at the end the frequency of the resulting run is indicated