Link performance prediction using spatial link performance mapping
In one embodiment, a current path of a mobile device is determined based on radio signals between the mobile device and a base station, which indicates a sequence of positions of the mobile device over a current time window. A future path of the mobile device is then predicted based on the current path, which indicates a sequence of predicted future positions of the mobile device over a future time window. A link performance prediction (LPP) is then generated for the mobile device based on the future path of the mobile device and a base station coverage map. The base station coverage map indicates a radio signal quality across a base station coverage area, which is represented as a three-dimensional (3D) coordinate space. Moreover, the LPP indicates a predicted performance of a radio link between the mobile device and the base station during the future time window.
1. A device for predicting radio link performance, comprising:
communication circuitry; and
processing circuitry to:
receive, via the communication circuitry, a link performance prediction (LPP) request from a requesting entity, wherein the LPP request is a request to predict a performance of a radio link between a mobile device and a base station during a future time window;
determine a current path of the mobile device, wherein the current path is determined based on radio signals between the mobile device and the base station, and wherein the current path indicates a sequence of positions of the mobile device over a current time window;
predict, based on the current path, a future path of the mobile device, wherein the future path indicates a sequence of predicted future positions of the mobile device over the future time window;
generate a link performance prediction (LPP) response for the mobile device based on the future path and a base station coverage map, wherein:
the base station coverage map indicates a radio signal quality across a base station coverage area of the base station, wherein the base station coverage area is represented in a three-dimensional (3D) coordinate space, wherein a plurality of 3D regions within the base station coverage area are represented as a plurality of points within the 3D coordinate space; and
the LPP response indicates a predicted performance of the radio link between the mobile device and the base station during the future time window; and
send, via the communication circuitry, the LPP response to the requesting entity.
2. The device of claim 1 , wherein the 3D coordinate space is based on:
a Cartesian coordinate system; or
a spherical coordinate system.
3. The device of claim 1 , wherein:
the current path of the mobile device comprises:
a first sequence of points within the 3D coordinate space, wherein the first sequence of points comprises a current point indicating a current position of the mobile device; and
a direction vector within the 3D coordinate space, wherein the direction vector indicates a current direction of travel of the mobile device;
the future path of the mobile device comprises a second sequence of points within the 3D coordinate space, wherein the second sequence of points comprises a next point indicating a next position of the mobile device, wherein the mobile device is predicted to transition from the current position to the next position; and
the processing circuitry to predict, based on the current path, the future path of the mobile device is further to:
predict, based on the first sequence of points and the direction vector, the second sequence of points.
4. The device of claim 3 , wherein the processing circuitry to predict, based on the first sequence of points and the direction vector, the second sequence of points is further to:
predict the second sequence of points using a predictive model, wherein the predictive model is trained to infer the second sequence of points from the first sequence of points and the direction vector, wherein the predictive model is trained based on a set of previous mobile device paths through the base station coverage area.
5. The device of claim 3 , wherein the processing circuitry to predict, based on the first sequence of points and the direction vector, the second sequence of points is further to:
determine, based on the current point and the direction vector, a set of potential transition points in the 3D coordinate space; and
predict the next point from the set of potential transition points using a predictive model, wherein the predictive model is trained to infer the next point from the current point and the set of potential transition points based on a set of previous mobile device paths through the base station coverage area.
6. The device of claim 5 , wherein the set of potential transition points comprises:
one or more adjacent points relative to the current point; and
one or more non-adjacent points relative to the current point, wherein the one or more non-adjacent points are identified based on one or more previous mobile device transitions from the current point to the one or more non-adjacent points.
7. The device of claim 1 , wherein the processing circuitry to generate the LPP response for the mobile device based on the future path and the base station coverage map is further to:
determine, based on the base station coverage map, a corresponding radio signal quality along the future path of the mobile device; and
generate the LPP response based on the corresponding radio signal quality along the future path of the mobile device.
8. The device of claim 1 , wherein:
the mobile device comprises a user equipment (UE) device; and
the base station comprises a millimeter wave (mmWave) base station associated with a radio access network (RAN).
9. At least one non-transitory machine-accessible storage medium having instructions stored thereon, wherein the instructions, when executed on processing circuitry, cause the processing circuitry to:
receive, via communication circuitry, a link performance prediction (LPP) request from a requesting entity, wherein the LPP request is a request to predict a performance of a radio link between a mobile device and a base station during a future time window;
determine a current path of the mobile device, wherein the current path is determined based on radio signals between the mobile device and the base station, and wherein the current path indicates a sequence of positions of the mobile device over a current time window;
predict, based on the current path, a future path of the mobile device, wherein the future path indicates a sequence of predicted future positions of the mobile device over the future time window;
generate a link performance prediction (LPP) response for the mobile device based on the future path and a base station coverage map, wherein:
the base station coverage map indicates a radio signal quality across a base station coverage area of the base station, wherein the base station coverage area is represented as a three-dimensional (3D) coordinate space; and
the LPP response indicates a predicted performance of the radio link between the mobile device and the base station during the future time window; and
send, via the communication circuitry, the LPP response to the requesting entity.
10. The at least one non-transitory machine-accessible storage medium of claim 9 , wherein: the base station coverage map represents a plurality of 3D regions within the base station coverage area as a plurality of points within the 3D coordinate space, wherein each point within the 3D coordinate space represents a corresponding 3D region within the base station coverage area; and the base station coverage map further indicates a corresponding radio signal quality at each of the plurality of 3D regions.
11. The at least one non-transitory machine-accessible storage medium of claim 10 , wherein the 3D coordinate space is based on: a Cartesian coordinate system; or a spherical coordinate system.
12. The at least one non-transitory machine-accessible storage medium of claim 10 , wherein: the current path of the mobile device further indicates: a current position of the mobile device; and a current direction of travel of the mobile device; and the future path of the mobile device further indicates: a next position of the mobile device, wherein the mobile device is predicted to transition from the current position to the next position.
13. The at least one non-transitory machine-accessible storage medium of claim 12 , wherein: the current path of the mobile device comprises: a first sequence of points within the 3D coordinate space, wherein the first sequence of points comprises a current point indicating the current position of the mobile device; and a direction vector within the 3D coordinate space, wherein the direction vector indicates the current direction of travel of the mobile device; the future path of the mobile device comprises a second sequence of points within the 3D coordinate space, wherein the second sequence of points comprises a next point indicating the next position of the mobile device; and the instructions that cause the processing circuitry to predict, based on the current path, the future path of the mobile device further cause the processing circuitry to: predict, based on the first sequence of points and the direction vector, the second sequence of points.
14. The at least one non-transitory machine-accessible storage medium of claim 13 , wherein the instructions that cause the processing circuitry to predict based on the first sequence of points and the direction vector, the second sequence of points further cause the processing circuitry to: predict the second sequence of points using a predictive model, wherein the predictive model is trained to infer the second sequence of points from the first sequence of points and the direction vector, wherein the predictive model is trained based on a set of previous mobile device paths through the base station coverage area.
15. The at least one non-transitory machine-accessible storage medium of claim 14 , wherein the predictive model comprises a recurrent neural network (RNN) model or a long short-term memory (LSTM) model.
16. The at least one non-transitory machine-accessible storage medium of claim 13 , wherein the instructions that cause the processing circuitry to predict based on the first sequence of points and the direction vector, the second sequence of points further cause the processing circuitry to: determine, based on the current point and the direction vector, a set of potential transition points in the 3D coordinate space; and predict the next point from the set of potential transition points, wherein the mobile device is predicted to transition from the current point to the next point.
17. The at least one non-transitory machine-accessible storage medium of claim 16 , wherein the set of potential transition points comprises: one or more adjacent points relative to the current point; and one or more non-adjacent points relative to the current point, wherein the one or more non-adjacent points are identified based on one or more previous mobile device transitions from the current point to the one or more non-adjacent points.
18. The at least one non-transitory machine-accessible storage medium of claim 16 , wherein the instructions that cause the processing circuitry to predict the next point from the set of potential transition points further cause the processing circuitry to: predict the next point using a predictive model, wherein the predictive model is trained to infer the next point from the current point and the set of potential transition points, wherein the predictive model is trained based on a set of previous mobile device paths through the base station coverage area.
19. The at least one non-transitory machine-accessible storage medium of claim 18 , wherein the predictive model comprises a Markov model.
20. The at least one non-transitory machine-accessible storage medium of claim 9 , wherein the instructions that cause the processing circuitry to generate the LPP response for the mobile device based on the future path and the base station coverage map further cause the processing circuitry to: determine, based on the base station coverage map, a corresponding radio signal quality along the future path of the mobile device; and generate the LPP response based on the corresponding radio signal quality along the future path of the mobile device.
21. The at least one non-transitory machine-accessible storage medium of claim 9 , wherein: the mobile device comprises a user equipment (UE) device; and the base station comprises a millimeter wave (mmWave) base station associated with a radio access network (RAN).
22. A method of predicting radio link performance, comprising:
receiving, via communication circuitry, a link performance prediction (LPP) request from a requesting entity, wherein the LPP request is a request to predict a performance of a radio link between a mobile device and a base station during a future time window;
determining a current path of the mobile device, wherein the current path is determined based on radio signals between the mobile device and the base station, and wherein the current path indicates a sequence of positions of the mobile device over a current time window;
predicting, based on the current path, a future path of the mobile device, wherein the future path indicates a sequence of predicted future positions of the mobile device over the future time window;
generating a link performance prediction (LPP) response for the mobile device based on the future path and a base station coverage map, wherein:
the base station coverage map indicates a radio signal quality across a base station coverage area of the base station, wherein the base station coverage area is represented in a three-dimensional (3D) coordinate space, wherein a plurality of 3D regions within the base station coverage area are represented as a plurality of points within the 3D coordinate space; and
the LPP response indicates a predicted performance of the radio link between the mobile device and the base station during the future time window; and
sending, via the communication circuitry, the LPP response to the requesting entity.
23. The method of claim 22 , wherein:
the current path of the mobile device comprises:
a first sequence of points within the 3D coordinate space, wherein the first sequence of points comprises a current point indicating a current position of the mobile device; and
a direction vector within the 3D coordinate space, wherein the direction vector indicates a current direction of travel of the mobile device;
the future path of the mobile device comprises a second sequence of points within the 3D coordinate space, wherein the second sequence of points comprises a next point indicating a next position of the mobile device, wherein the mobile device is predicted to transition from the current position to the next position; and
predicting, based on the current path, the future path of the mobile device comprises:
predicting, based on the first sequence of points and the direction vector, the second sequence of points.
24. The method of claim 23 , wherein predicting, based on the first sequence of points and the direction vector, the second sequence of points comprises:
determining, based on the current point and the direction vector, a set of potential transition points in the 3D coordinate space, wherein the set of potential transition points comprises:
one or more adjacent points relative to the current point; and
one or more non-adjacent points relative to the current point, wherein the one or more non-adjacent points are identified based on one or more previous mobile device transitions from the current point to the one or more non-adjacent points; and
predicting the next point from the set of potential transition points, wherein the mobile device is predicted to transition from the current point to the next point.
25. The method of claim 22 , wherein:
the mobile device comprises a user equipment (UE) device; and
the base station comprises a millimeter wave (mmWave) base station associated with a radio access network (RAN).