Region-adapted neural network for location determination
A wireless device includes a satellite receiver to receive data from multiple satellites. The wireless device also includes processing circuitry and memory. The memory stores one or more neural network models. The processing circuitry is operative to identify a neural network model that has been trained to adapt to a region in which the wireless device operates, classify satellite raw measurements from each satellite at a given time into a corresponding quality level using the neural network model, and identify satellite raw measurements with a quality level higher than a threshold. The location of the wireless device is calculated using the identified satellite raw measurements.
1. A method for determining a location of a wireless device that receives data from a plurality of satellites, comprising:
selecting, from a plurality of neural network models adapting to respective regions, a neural network model that has been trained to adapt to a region encompassing the location of the wireless device;
classifying, by using the neural network model, satellite raw measurements from each satellite at a given time into a corresponding quality level; and
calculating the location of the wireless device using the satellite raw measurements that are classified as having a quality level higher than a threshold.
2. The method of claim 1 , wherein the neural network model has been trained to classify the satellite raw measurements according to a pseudo-range error computed during a training phase, the pseudo-range error including at least a multipath propagation error.
3. The method of claim 1 , wherein the neural network model is a multi-layer perceptron network.
4. The method of claim 1 , wherein the satellite raw measurements include some or all of satellite raw measurement items: a satellite system identifier, signal-to-noise ratio (SNR), Doppler information, Elevation, Azimuth, and clock information.
5. The method of claim 4 , wherein the neural network model includes a plurality of input nodes at an input layer, each input node receiving a respective one of the satellite raw measurement items.
6. The method of claim 1 , further comprising:
disregarding the satellite raw measurements having the quality level lower than the threshold.
7. The method of claim 1 , further comprising:
estimating the location of the wireless device to determine whether the neural network model stored in a device memory corresponds to the estimated location; and
downloading, from a server, a new neural network model that corresponds to the estimated location when the neural network model stored in the device memory does not correspond to the estimated location.
8. The method of claim 1 , further comprising:
incorporating building information to estimate multipath effects on signals from a given satellite at a given time; and
classifying the satellite raw measurements based, at least in part, on the building information.
9. The method of claim 1 , further comprising:
incorporating building information to estimate multipath effects on signals from a given satellite at a given time; and
calculating the location of the wireless device using the identified satellite raw measurements and the building information.
10. The method of claim 1 , further comprising:
incorporating information from one or more additional wireless systems in the calculation of the location of the wireless device, the one or more additional wireless systems include one or more of: a cellular system, a Wi-Fi system, and a Bluetooth system.
11. A wireless device, comprising:
a satellite receiver operative to receive data from a plurality of satellites;
processing circuitry; and
memory to store a plurality of neural network models adapting to respective regions, wherein the processing circuitry is operative to:
select, from the plurality of neural network models, a neural network model that has been trained to adapt to a region encompassing the location of the wireless device;
classify, by using the neural network model, satellite raw measurements from each satellite at a given time into a corresponding quality level using the neural network model; and
calculate the location of the wireless device using the satellite raw measurements that are classified as having a quality level higher than a threshold.
12. The wireless device of claim 11 , wherein the neural network model has been trained to classify the satellite raw measurements according to a pseudo-range error computed during a training phase, the pseudo-range error including at least a multipath propagation error.
13. The wireless device of claim 11 , wherein the neural network model is a multi-layer perceptron network.
14. The wireless device of claim 11 , wherein the satellite raw measurements include some or all of satellite raw measurement items: a satellite system identifier, signal-to-noise ratio (SNR), Doppler information, Elevation, Azimuth, and clock information.
15. The wireless device of claim 14 , wherein the neural network model includes a plurality of input nodes at an input layer, each input node receiving a respective one of the satellite raw measurement items.
16. The wireless device of claim 11 , wherein when identifying the satellite raw measurements, the processing circuitry is further operative to:
disregard the satellite raw measurements having the quality level lower than the threshold.
17. The wireless device of claim 11 , wherein the processing circuitry is further operative to:
estimate the location of the wireless device to determine whether the neural network model stored in the memory corresponds to the estimated location; and
download, from a server, a new neural network model that corresponds to the estimated location when the neural network model stored in the memory does not correspond to the estimated location.
18. The wireless device of claim 11 , wherein when classifying the satellite raw measurements, the processing circuitry is further operative to:
incorporate building information to estimate multipath effects on signals from a given satellite at a given time; and
classify the satellite raw measurements based, at least in part, on the building information.
19. The wireless device of claim 11 , wherein when calculating the location of the wireless device, the processing circuitry is further operative to:
incorporate building information to estimate multipath effects on signals from a given satellite at a given time; and
calculate the location of the wireless device using the identified satellite raw measurements and the building information.
20. The wireless device of claim 11 , wherein when calculating the location of the wireless device, the processing circuitry is further operative to:
incorporate information from one or more additional wireless systems in the calculation of the location of the wireless device, the one or more additional wireless systems include one or more of: a cellular system, a Wi-Fi system, and a Bluetooth system.