android module

Functions to process Android measurements.

Tested on Google Android’s GNSSLogger App v3.0.6.4

Includes functionality for:

Fix Measurements
Raw Measurements
Accel Measurements
Gyro Measurements
Mag Measurements
Bearing Measurements

class android.AndroidRawAccel(input_path, sensor_fields=('UncalAccel', 'Accel'))[source]

Bases: NavData

Class handling Accelerometer measurements from Android.

Inherits from NavData().

Parameters:

input_path (string or path-like) – File location of data file to read.
sensor_fields (tuple) – Names for the sensors to extract from the full log file.

_row_map()[source]

Map of row names from loaded to gnss_lib_py standard

Returns:: row_map – Dictionary of the form {old_name : new_name}
Return type:: Dict

postprocess()[source]: Postprocess loaded data.

preprocess(input_path)[source]

Read Android raw file and produce Accel dataframe objects.

Parameters:: input_path (string or path-like) – File location of data file to read.
Returns:: measurements – Dataframe that contains the accel measurements from the log.
Return type:: pd.DataFrame

class android.AndroidRawFixes(input_path)[source]

Bases: NavData

Class handling location fix measurements from an Android log.

Inherits from NavData().

Parameters:: input_path (string or path-like) – Path to measurement csv or txt file.

static _provider_map()[source]

Map to more intuitive names for fix provider.

Returns:: provider_map – Dictionary of the form {old_name : new_name}
Return type:: Dict

static _row_map()[source]

Map of row names from loaded to gnss_lib_py standard

Returns:: row_map – Dictionary of the form {old_name : new_name}
Return type:: Dict

postprocess()[source]: Postprocess loaded data.

preprocess(input_path)[source]

Read Android raw file and produce location fix dataframe objects

Parameters:: input_path (string or path-like) – File location of data file to read.
Returns:: fix_df – Dataframe that contains the location fixes from the log.
Return type:: pd.DataFrame

class android.AndroidRawGnss(input_path, filter_measurements=True, measurement_filters={'adr_uncertainty': 15.0, 'adr_valid': True, 'arrival_time': True, 'bias_uncertainty': 40.0, 'bias_valid': True, 'state_decoded': True, 'sv_time_uncertainty': 500.0, 'time_valid': True, 'unknown_constellations': True}, remove_rx_b_from_pr=False, verbose=False)[source]

Bases: NavData

Handles Raw GNSS measurements from Android.

Data types in the Android’s GNSSStatus messages are documented on their website [1].

Parameters:

input_path (string or path-like) – Path to measurement csv or txt file.
filter_measurements (bool) – Filter noisy measurements based on known conditions.
measurement_filters (dict) – Conditions under which measurements should be filtered. An emptry dictionary passed into measurement_filters is equivalent to setting filter_measurements to False. See the docstring for filter_raw_measurements for details.
remove_rx_b_from_pr (bool) – If true, removes the estimated initial receiver clock bias from the pseudorange values.
verbose (bool) – If true, prints extra debugging statements.

References

static _row_map()[source]

Map of row names from loaded to gnss_lib_py standard

Returns:: row_map – Dictionary of the form {old_name : new_name}
Return type:: Dict

filter_raw_measurements(t_rx_secs)[source]

Filter noisy measurements.

Filter conditions are set in the AndroidRawGnss() initialization with the measurement_filters variable. The general framework for measurement filters is taken from Google’s ION GNSS+ 2023 workshop presentation [4]. Google’s gps-measurement-tools opensource software shows a couple of their implementations with example values [5] [6].

The possible keys to include in the measurement_filters dictionary variable include:

bias_valid : If true, measurements where FullBiasNanos is greater or equal to zero will be removed.

bias_uncertainty : Any measurements will be re where BiasUncertaintyNanos is greater than the set value.

arrival_time : If true, measurements where the arrival time (t_rx_secs) is negative or unrealistically large will be removed.

unknown_constellations : If true, measurements with an unknown_constellation will be removed.

time_valid : If true, measurements with where TimeNanos is empty will be removed.

state_decoded : If true, will filter measurements that don’t have the time-of-week decoded [7].

sv_time_uncertainty : Any measurements will be removed where ReceivedSvTimeUncertaintyNanos is greater than the set value.

adr_valid : If true, measurements where AccumulatedDeltaRangeState is not valid will be removed.

adr_uncertainty : Any measurements will be removed where ReceivedSvTimeUncertaintyNanos is greater than the set value.

Only keys present in the measurement_filters dictionary will be used for the filter. For example if ``measurement_filters = {``bias_valid : True}, then only measurements will invalid FullBiasNanos will be filtered and no other measurements will be filtered.

Parameters:: t_rx_secs (np.ndarray) – arrival time computed in AndroidRawGnss.postprocess().

References

postprocess()[source]

Postprocess loaded NavData.

Strategy for computing the arrival time was taken from an EUSPA white paper [2] and Google’s source code Arrival time taken from [3].

References

preprocess(input_path)[source]

Extract Raw measurements.

Built on the first parts of make_gnss_dataframe and correct_log

Parameters:: input_path (string or path-like) – Path to measurement csv or txt file.
Returns:: measurements – Subset of “Raw” measurements.
Return type:: pd.DataFrame