IP Library Granted Patent US 12,652,578
Granted Patent B2
US 12,652,578 · App. 18/104,081 · Granted Jun 9, 2026

Preventing service interruptions by predicting outages in a satellite network

Inventors: Thomas Szigeti (Vancouver, CA); David John Zacks (Vancouver, CA); Jeff Apcar (Willoughby, AU); Robert Edgar Barton (Richmond, CA)
Assignee: Cisco Technology, Inc.
H04W28/0942H04W16/22H04W76/18
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,652,578
App. No.
18/104,081
Granted
Jun 9, 2026
Kind
B2
Abstract

This disclosure describes techniques for predicting and accommodating for outages in a satellite network using crowdsourced data. An example method includes receiving outage data indicating first outages experienced by first endpoints in a first geographical region. The first outages, for instance, include interruptions in communication between first satellites and the first endpoints. The example method further includes predicting, based on the outage data, a second outage comprising an interruption in communication between at least one second satellite and a second endpoint in a second geographical region. Further, the example method includes causing the second endpoint to transmit user data over a secondary network in advance of the second outage.

Claims (52)

1 . A method performed by a system, the method comprising:

receiving outage data indicating first outages experienced by first endpoints in a first geographical region, the first outages comprising interruptions in communication between first satellites and the first endpoints;

receiving contextual data indicating:

radio frequency (RF) conditions impacting the first geographical region and a second geographical region;

weather conditions in the first geographical region and the second geographical region; and

network conditions associated with a satellite network comprising the first satellites and at least one second satellite;

predicting, based on the outage data and the contextual data, a future time associated with a second outage comprising an interruption in communication between the at least one second satellite and a second endpoint in a second geographical region; and

causing the second endpoint to transmit user data over a secondary network in advance of the future time associated with the second outage.

2 . The method of claim 1 , wherein the outage data is received from at least one proxy connected to the first endpoints.

3 . The method of claim 1 , wherein the outage data comprises locations of the first endpoints during the first outages, times of the first outages, and durations of the first outages.

4 . The method of claim 1 , wherein the first satellites comprise at least one Low Earth Orbit (LEO) satellite.

5 . The method of claim 1 , wherein predicting the second outage comprises:

training a machine learning model using a first portion of the outage data; and

in response to training the machine learning model, predicting the second outage based on a second portion of the outage data and the machine learning model.

6 . The method of claim 1 , wherein the secondary network comprises a terrestrial cellular network.

7 . The method of claim 1 , wherein causing the second endpoint to transmit user data over the secondary network in advance of the second outage comprises:

transmitting, to a proxy connected to the second endpoint, an instruction to transmit the user data to the secondary network.

8 . The method of claim 1 , wherein the contextual data further indicates

meteorological events impacting the first geographical region and the second geographical region; and

at least one inoperative satellite among the first satellites and/or at least one second satellite.

9 . The method of claim 1 , wherein the contextual data further indicates meteorological events impacting the first geographical region and the second geographical region the meteorological events comprising solar flares.

10 . A system, comprising

at least one processor; and

one or more non-transitory media storing instructions that, when executed by the system, cause the system to perform operations comprising:

receiving outage data indicating first outages experienced by first endpoints in a first geographical region, the first outages comprising interruptions in communication between first satellites and the first endpoints;

receiving contextual data indicating radio frequency (RF) conditions impacting the first geographical region and a second geographical region, weather conditions in the first geographical region and the second geographical region, or network conditions associated with a satellite network comprising the first satellites and at least one second satellite;

predicting, based on the outage data and the contextual data, a second outage comprising an interruption in communication between the at least one second satellite and a second endpoint in a second geographical region; and

in response to predicting the second outage and prior to the second outage, transmitting, to a proxy communicatively coupled to the second endpoint, a message indicating the second outage.

11 . The system of claim 10 , wherein the outage data is received from at least one proxy connected to the first endpoints.

12 . The system of claim 10 , wherein the outage data comprises locations of the first endpoints during the first outages, times of the first outages, and durations of the first outages.

13 . The system of claim 10 , wherein the first satellites comprise at least one Low Earth Orbit (LEO) satellite.

14 . The system of claim 10 , wherein predicting the second outage comprises:

training a machine learning model using a first portion of the outage data; and

in response to training the machine learning model, predicting the second outage based on a second portion of the outage data and the machine learning model.

15 . The system of claim 10 , wherein the message indicating the second outage comprises an instruction to transmit user data associated with the second endpoint over a terrestrial network, rather than the satellite network.

16 . The system of claim 15 , wherein the terrestrial network comprises a cellular network.

17 . An outage prediction system, comprising:

at least one processor; and

memory storing instructions that, when executed by the at least one processor, cause the at least one processor to perform operations comprising:

receiving outage data indicating first outages experienced by endpoints disposed in a grid of geographical regions, the first outages comprising interruptions in communication between a satellite network and the endpoints, the satellite network comprising low earth orbit (LEO) satellites;

receiving contextual data indicating radio frequency (RF) conditions in the geographical regions, weather conditions in the geographical regions, and network conditions associated with the satellite network;

predicting, based on the outage data and the contextual data, a second outage comprising an interruption in communication between the satellite network and a particular endpoint among the endpoints, the particular endpoint being in one of the geographical regions; and

in response to predicting the second outage, transmitting, to a proxy communicatively coupled to the particular endpoint, an instruction to route data associated with the particular endpoint over a terrestrial cellular network in advance of the second outage.

18 . The system of claim 17 , wherein the first outages are experienced by at least one of the endpoints in a first geographical region among the geographical regions,

wherein the second outage is predicted to impact the particular endpoint in a second geographical region among the geographical regions, and

wherein the first geographical region borders the second geographical region.

19 . The system of claim 17 , wherein predicting, based on the outage data and the contextual data, the second outage comprising the interruption in communication between the satellite network and the particular endpoint comprises:

training a machine learning (ML) model based on a first portion of the outage data and a first portion of the contextual data;

in response to training the ML model, predicting the second outage based on the ML model, a second portion of the outage data, and a second portion of the contextual data.

20 . The system of claim 17 , wherein the first outages and the second outage are caused by at least one of:

an inoperative LEO satellite among the LEO satellites; or

a weather pattern moving across the geographical regions.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jan 31, 2023
From: SZIGETI, THOMAS; ZACKS, DAVID JOHN; APCAR, JEFF; BARTON, ROBERT EDGAR
To: CISCO TECHNOLOGY, INC.
Reel/Frame 062553/0004 →
Continuity (1)
Related Publication 20240259875A1 · Aug 1, 2024
References Cited (12)
US 6070073A · Maveddat et al. · 2000 [cited by applicant]
US 9297723B1 · Hofmann et al. · 2016 [cited by applicant]
US 11044008B2 · Ravishankar · 2021 [cited by examiner]
US 20160198152A1 · Chin · 2016 [cited by examiner]
US 20210399972A1 · Nguyen et al. · 2021 [cited by applicant]
US 20220209876A1 · Sitaram et al. · 2022 [cited by applicant]
US 20220239367A1 · Mayer · 2022 [cited by examiner]
US 20220360323A1 · Dooling · 2022 [cited by examiner]
US 20230388774A1 · Vivanco · 2023 [cited by examiner]
US 20240172302A1 · Rakshit · 2024 [cited by examiner]
US 20240187090A1 · Wu · 2024 [cited by examiner]
Hassanien, et al., “Machine Learning in Telemetry Data Mining of Space Mission: Basics, Challenging and Future Directions,” Artificial Intelligence Review, SpringerLink, Sep. 4, 2019, 30 pages. [cited by applicant]