Rogue base station router detection with machine learning algorithms

1. A method for determining if a cellular attack from a rogue base station router (RBSR) on a cellular network is active, the method comprising:

determining a set of broadcast data associated with a plurality of base station routers (BSRs) operating in the cellular network;

determining, via inputting the set of broadcast data to a machine learning algorithm, anomalous broadcast data in the set of broadcast data;

identifying a BSR associated with the anomalous broadcast data as the RBSR;

correlating the identification of the RBSR with an event associated with ft the cellular attack in the cellular network; and

determining that the cellular attack from the RBSR on the cellular network is active based on the correlating the identification of the RBSR with the event.

2. The method of claim 1 , wherein the set of broadcast data comprises, for each BSR of the plurality of BSRs, featurized system information blocks (SIBs) associated with that BSR.

3. The method of claim 1 , wherein the anomalous broadcast data associated with the RBSR does not include one or more predetermined parameters found in broadcast data associated with legitimate BSRs of the plurality of BSRs operating in the cellular network.

4. The method of claim 1 , further comprising:

determining a confidence level that the cellular attack from the RBSR on the cellular network is active based on the correlating the identification of the RBSR with the event; and

generating, based on the confidence level exceeding a predetermined threshold, notifications for users on the cellular network, wherein the notifications indicate that the cellular attack from the RBSR is active.

5. A method for determining if a cellular attack from a rogue base station router (RBSR) on a cellular network is active, the method comprising:

determining a set of broadcast data associated with a plurality of base station routers (BSRs) operating in the cellular network;

inserting broadcast data associated with a known RBSR into the set of broadcast data;

inputting the set of broadcast data to a machine learning algorithm;

determining that the machine learning algorithm similarly classifies the RBSR and the known RBSR;

determining that an event associated with the cellular attack is present in the cellular network; and

determining that the cellular attack from the RBSR on the cellular network is active based on the determining the event and the similar classification of the RBSR and the known RBSR by the machine learning algorithm.

6. The method of claim 5 , wherein the machine learning algorithm is configured to classify an input BSR as, at the least, legitimate or rogue.

7. The method of claim 5 , wherein the set of broadcast data is determined by a single scan of the cellular network.

8. The method of claim 5 , wherein the machine learning algorithm comprises an unsupervised machine learning algorithm.

9. The method of claim 8 , wherein the machine learning algorithm comprises a clustering machine learning algorithm, and

the determining that the machine learning algorithm similarly classifies the RBSR and the known RBSR comprises determining that the machine learning algorithm assigns the RBSR and the known RBSR to a same cluster.

10. A non-transition computer readable medium including program instructions which when executed by a processor are configured to: determine a set of broadcast data associated with a plurality of base station routers (BSRs) operating in a cellular network; determine, via inputting the set of broadcast data to a machine learning algorithm, anomalous broadcast data in the set of broadcast data; identify a BSR associated with the anomalous broadcast data as a rogue BSR (RBSR); correlate the identification of the RBSR with an event associated with a cellular attack in the cellular network; and determine that the cellular attack from the RBSR on the cellular network is active based on the correlation of the RBSR with the event.

11. The computer readable medium of claim 10 , wherein the set of broadcast data comprises, for each BSR of the plurality of BSRs, featurized system information blocks (SIBs) associated with that BSR.

12. The computer readable medium of claim 10 , wherein the anomalous broadcast data associated with the RBSR does not include one or more predetermined parameters found in broadcast data associated with legitimate BSRs of the plurality of BSRs operating in the cellular network.

13. The computer readable medium of claim 10 , wherein the program instructions which when executed by the processor are further configured to:

determine a confidence level that the cellular attack from the RBSR on the cellular network is active based on the correlating the identification of the RBSR with the event; and

generate, based on the confidence level exceeding a predetermined threshold, notifications for users on the cellular network, wherein the notifications indicate that the cellular attack from the RBSR is active.

14. A non-transition computer readable medium including program instructions which when executed by a processor are configured to: determine a set of broadcast data associated with a plurality of base station routers (BSRs) operating in a cellular network; insert broadcast data associated with a known rogue RBSR (RBSR) into the set of broadcast data; input the set of broadcast data to a machine learning algorithm; determine the machine learning algorithm similarly classifies the RBSR and the known RBSR; determine an event associated with a cellular attack is present in the cellular network; and determine the cellular attack from the RBSR on the cellular network is active based on the determined event and the similar classification of the RBSR and the known RBSR by the machine learning algorithm.

15. The computer readable medium of claim 14 , wherein the machine learning algorithm is configured to classify an input BSR as, at the least, legitimate or rogue.

16. The computer readable medium of claim 14 , wherein the set of broadcast data is determined by a single scan of the cellular network.

17. The computer readable medium of claim 14 , wherein the machine learning algorithm comprises an unsupervised machine learning algorithm.

18. The computer readable medium of claim 17 , wherein the machine learning algorithm comprises a clustering machine learning algorithm, and

the determining the machine learning algorithm program instructions similarly classifies the RBSR and the known RBSR determines the machine learning algorithm assigns the RBSR and the known RBSR to a same cluster.