low-power gps tracking software - eth z · 2018-05-04 · low-power gps tracking software low-power...

1
Distributed Computing Prof. R. Wattenhofer BA/MA/SA/Group/Lab: Low-Power GPS Tracking Software Low-power devices for the Internet of Things (IoT) have very little energy and computation power available. Still, they should provide a lot of functionality to enable “smart” applica- tions. For instance, one would always like to know where such a device is located. We try to tackle this problem using GPS, but with an approach very different from current receivers. At our group, we developed an al- gorithm for a low-power GPS receiver. The receiver only samples the signal and then sends the samples to a server, which then does the expensive processing to get the position. Of course, we now have a heavy computational load on the server. The goal of this project is to reduce the computation power at the server. We do have many ideas how this can be achieved. For example by keeping track of the receiver state over time, thus having to update the position only incrementally at each measurement point. You will implement these improvements in our existing implementation, which uses newest generation GPUs to process the recorded signals. Requirements: Creative thinking and advanced programming skills are advantageous to successfully work on this topic. Experience with GPU programming is a plus. The student(s) should be able to work independently! Interested? Please contact us for more details! Contact Pascal Bissig: [email protected], ETZ G95 Manuel Eichelberger: [email protected], ETZ G97

Upload: others

Post on 06-Aug-2020

3 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Low-Power GPS Tracking Software - ETH Z · 2018-05-04 · Low-Power GPS Tracking Software Low-power devices for the Internet of Things (IoT) have very little energy and computation

Distributed Computing

Prof. R. WattenhoferBA/MA/SA/Group/Lab:

Low-Power GPS Tracking Software

Low-power devices for the Internet of Things (IoT) have very little energy and computationpower available. Still, they should provide a lot of functionality to enable “smart” applica-tions. For instance, one would always like to know where such a device is located. We tryto tackle this problem using GPS, but with an approach very different from current receivers.

At our group, we developed an al-gorithm for a low-power GPS receiver.The receiver only samples the signal andthen sends the samples to a server, whichthen does the expensive processing toget the position. Of course, we nowhave a heavy computational load on theserver.

The goal of this project is to reduce thecomputation power at the server. We do havemany ideas how this can be achieved. Forexample by keeping track of the receiver state over time, thus having to update the positiononly incrementally at each measurement point. You will implement these improvements inour existing implementation, which uses newest generation GPUs to process the recordedsignals.

Requirements: Creative thinking and advanced programming skills are advantageous tosuccessfully work on this topic. Experience with GPU programming is a plus. The student(s)should be able to work independently!

Interested? Please contact us for more details!

Contact

• Pascal Bissig: [email protected], ETZ G95

• Manuel Eichelberger: [email protected], ETZ G97