Systems and methods for vehicular-network-assisted federated machine learning
Systems and methods for vehicular-network-assisted federated machine learning are disclosed herein. One embodiment transmits first metadata from a connected vehicle to at least one other connected vehicle; receives, at the connected vehicle, second metadata from the at least one other connected vehicle; receives, at the connected vehicle based on analysis of the first and second metadata, a notification that the connected vehicle has been elected to participate in the current training phase of a federated machine learning process; receives, at the connected vehicle, instructions to prepare the connected vehicle for the next training phase; trains a machine learning model to perform a task at the connected vehicle during the current training phase to produce a locally trained machine learning model; and submits the locally trained machine learning model for aggregation with at least one other locally trained machine learning model to produce an aggregated locally trained machine learning model.
1. A system for vehicular-network-assisted federated machine learning, the system comprising:
a processor; and
a memory storing computer-readable instructions that, when executed by the processor, cause the processor to:
exchange metadata between a connected vehicle and at least one other connected vehicle in a vehicular micro cloud, wherein the metadata includes information regarding sensor capabilities of the connected vehicle and the at least one other connected vehicle;
receive, at the connected vehicle based on an analysis of the metadata, a notification that the connected vehicle has been elected to participate in a current training phase of a federated machine learning process;
receive, at the connected vehicle, instructions that position the connected vehicle to provide information about an object of interest to the federated machine learning process;
train a machine learning model to perform a task at the connected vehicle during the current training phase to produce a locally trained machine learning model; and
submit the locally trained machine learning model for aggregation with at least one other locally trained machine learning model produced by at least one other elected vehicle in the vehicular micro cloud to produce an aggregated locally trained machine learning model that is used as a starting point for a next training phase of the federated machine learning process.
2. The system of claim 1 , wherein the metadata includes information regarding at least one of availability of particular types of sensors, capabilities of specific available sensors, quality of the specific available sensors, and an itinerary that includes scheduled stops.
3. The system of claim 1 , wherein the connected vehicle receives the instructions that position the connected vehicle through coordination among a cloud leader of the vehicular micro cloud and at least one other cloud leader of a different vehicular micro cloud within a same vehicular macro cloud.
4. The system of claim 1 , wherein the analysis of the metadata is performed at one of a cloud server and an edge server, the analysis of the metadata resulting in the connected vehicle being elected to participate in the current training phase of the federated machine learning process.
5. The system of claim 1 , wherein the analysis of the metadata is performed at the connected vehicle and the at least one other connected vehicle in a distributed fashion, the analysis of the metadata resulting in the connected vehicle being elected to participate in the current training phase of the federated machine learning process.
6. The system of claim 1 , wherein the connected vehicle and the at least one other connected vehicle communicate via one or more vehicle-to-vehicle (V2V) communication links.
7. The system of claim 1 , wherein the aggregated locally trained machine learning model is further aggregated with at least one other aggregated locally trained machine learning model from another vehicular micro cloud at one of a cloud server and an edge server in preparation for the next training phase of the federated machine learning process.
8. The system of claim 1 , wherein the connected vehicle is an autonomous vehicle.
9. A non-transitory computer-readable medium for vehicular-network-assisted federated machine learning and storing instructions that, when executed by a processor, cause the processor to:
exchange metadata between a connected vehicle and at least one other connected vehicle in a vehicular micro cloud, wherein the metadata includes information regarding sensor capabilities of the connected vehicle and the at least one other connected vehicle;
receive, at the connected vehicle based on an analysis of the metadata, a notification that the connected vehicle has been elected to participate in a current training phase of a federated machine learning process;
receive, at the connected vehicle, instructions that position the connected vehicle to provide information about an object of interest to the federated machine learning process;
train a machine learning model to perform a task at the connected vehicle during the current training phase to produce a locally trained machine learning model; and
submit the locally trained machine learning model for aggregation with at least one other locally trained machine learning model produced by at least one other elected vehicle in the vehicular micro cloud to produce an aggregated locally trained machine learning model that is used as a starting point for a next training phase of the federated machine learning process.
10. The non-transitory computer-readable medium of claim 9 , wherein the metadata includes information regarding at least one of availability of particular types of sensors, capabilities of specific available sensors, quality of the specific available sensors, and an itinerary that includes scheduled stops.
11. A method, comprising:
exchanging metadata between a connected vehicle and at least one other connected vehicle in a vehicular micro cloud, wherein the metadata includes information regarding sensor capabilities of the connected vehicle and the at least one other connected vehicle;
receiving, at the connected vehicle based on an analysis of the metadata, a notification that the connected vehicle has been elected to participate in a current training phase of a federated machine learning process;
receiving, at the connected vehicle, instructions that position the connected vehicle to provide information about an object of interest to the federated machine learning process;
training a machine learning model to perform a task at the connected vehicle during the current training phase to produce a locally trained machine learning model; and
submitting the locally trained machine learning model for aggregation with at least one other locally trained machine learning model produced by at least one other elected vehicle in the vehicular micro cloud to produce an aggregated locally trained machine learning model that is used as a starting point for a next training phase of the federated machine learning process.
12. The method of claim 11 , wherein the metadata includes information regarding at least one of availability of particular types of sensors, capabilities of specific available sensors, quality of the specific available sensors, and an itinerary that includes scheduled stops.
13. The method of claim 11 , wherein the analysis of the metadata is performed at one of a cloud server and an edge server, the analysis of the metadata resulting in the connected vehicle being elected to participate in the current training phase of the federated machine learning process.
14. The method of claim 11 , wherein the analysis of the metadata is performed at the connected vehicle and the at least one other connected vehicle in a distributed fashion, the analysis of the first and second metadata resulting in the connected vehicle being elected to participate in the current training phase of the federated machine learning process.
15. The method of claim 11 , wherein the connected vehicle and the at least one other connected vehicle communicate via one or more vehicle-to-vehicle (V2V) communication links.
16. The method of claim 11 , wherein the aggregated locally trained machine learning model is further aggregated with at least one other aggregated locally trained machine learning model from another vehicular micro cloud at one of a cloud server and an edge server in preparation for the next training phase of the federated machine learning process.