Methods, systems, and computer readable media for 5G user equipment (UE) historical mobility tracking and security screening using mobility patterns

Abstract

A method for historical 5G user equipment (UE) mobility tracking and security screening includes receiving, at a network data aggregation node including at least one processor, UE registration data from 5G network functions (NFs) as UEs connect to different network locations. The method further includes aggregating, at the network node, registration data for individual UEs from the 5G NFs to produce mobility patterns for the UEs. The method further includes receiving, at the network node and from a 5G NF located in a home network of a UE, a request for a mobility pattern of the UE in response to receiving a message for effecting a new registration for the UE. The method further includes responding to the request by transmitting the mobility pattern to the 5G NF located in the home network of the UE.

Claims

1. A method for historical 5G user equipment (UE) mobility tracking and security screening, the method comprising: receiving, at a network data aggregation node including at least one processor, UE registration data from 5G network functions (NFs) as UEs connect to different network locations; aggregating, at the network node, registration data for individual UEs to produce mobility patterns for the UEs, wherein aggregating the registration data to generate the mobility patterns comprises storing, for each mobility pattern, an indicator of location of a UE, a timestamp of when the UE registered at the location, and a type allocation code (TAC) for each instance of registration of the UE; receiving, at the network node and from a 5G NF located in a home network of a UE, a request for the mobility pattern of the UE in response to receiving a message for effecting a new registration for the UE; responding to the request by transmitting the mobility pattern to the 5G NF located in the home network of the UE, wherein the 5G NF located in the home network of the UE comprises a security edge protection proxy (SEPP) and further comprising, at the SEPP, receiving the mobility pattern and determining, based on the mobility pattern, that the message for effecting the new registration indicates a UE registration pattern anomaly, wherein determining that the message for effecting the new registration indicates a UE registration pattern anomaly comprises: determining a device type from the TAC; comparing the mobility pattern to mobility patterns of devices of the same or similar type as the UE; and determining that the registration is anomalous based on results of the comparison; and at the SEPP, in response to determining that the message for effecting the new registration indicates a UE registration pattern anomaly, blocking the message for effecting the new registration.

2. The method of claim 1 wherein receiving the UE registration data includes receiving a mobility pattern from at least one of the 5G NFs.

3. The method of claim 1 wherein the network data aggregation node comprises a network data analytics platform (NWDAF) or a unified data repository (UDR).

4. The method of claim 1 wherein the network data aggregation node comprises a non-3GPP defined network data aggregation platform.

5. The method of claim 1 wherein the UE comprises an Internet of things (IoT) device.

6. A system for historical 5G user equipment (UE) mobility tracking and security screening, the system comprising: a network data aggregation node including at least one processor; and 5G UE mobility pattern generator for receiving registration data from 5G network functions (NFs) as UEs connect to different network locations, aggregating registration data for individual UEs to produce mobility patterns for the UEs, wherein aggregating the registration data to generate the mobility patterns comprises storing, for each mobility pattern, an indicator of location of a UE, a timestamp of when the UE registered at the location, and a type allocation code (TAC) for each instance of registration of the UE, wherein the 5G UE mobility pattern generator is configured for receiving, from a 5G NF located in a home network of a UE, a request for a mobility pattern of the UE in response to receiving a message for effecting a new registration for the UE, and responding to the request by transmitting the mobility pattern to the home network of the UE, wherein the 5G NF located in the home network of the UE comprises a security edge protection proxy (SEPP) and the SEPP is configured to: receive the mobility pattern; determining, based on the mobility pattern, that the message for effecting the new registration indicates a UE registration pattern anomaly, wherein determining that the message for effecting the new registration indicates a UE registration pattern anomaly comprises determining a device type from the TAC, comparing the mobility pattern to mobility patterns of devices of the same or similar type as the UE, and determining that the registration is anomalous based on results of the comparison; and in response to determining that the message for effecting the new registration indicates a UE registration pattern anomaly, block the message for effecting the new registration.

7. The system of claim 6 wherein receiving the UE registration data includes receiving a mobility pattern for the UE from at least one of the 5G NFs.

8. The system of claim 6 wherein the network data aggregation node comprises a network data analytics platform (NWDAF) or a unified data repository (UDR).

9. The system of claim 6 wherein the network data aggregation node comprises a non-3GPP defined network data aggregation platform.

10. A non-transitory computer readable medium having stored thereon executable instructions that when executed by a processor of a computer control the computer to perform steps comprising: receiving, at a network data aggregation node including at least one processor, UE registration data from 5G network functions (NFs) as UEs connect to different network locations; aggregating, at the network node, registration data for individual UEs from the 5G NFs to produce mobility patterns for the UEs, wherein aggregating the registration data to generate the mobility patterns comprises storing, for each mobility pattern, an indicator of location of a UE, a timestamp of when the UE registered at the location, and a type allocation code (TAC) for each instance of registration of the UE; receiving, at the network node and from a 5G NF located in a home network of a UE, a request for the mobility pattern of the UE in response to receiving a message for effecting a new registration for the UE; and responding to the request by transmitting the mobility pattern to the 5G NF located in the home network of the UE, wherein the 5G NF located in the home network of the UE comprises a security edge protection proxy (SEPP) and further comprising, at the SEPP, receiving the mobility pattern and determining, based on the mobility pattern, that the message for effecting the new registration indicates a UE registration pattern anomaly, wherein determining that the message for effecting the new registration indicates a UE registration pattern anomaly comprises: determining a device type from the TAC; comparing the mobility pattern to mobility patterns of devices of the same or similar type as the UE; and determining that the registration is anomalous based on results of the comparison; and at the SEPP, in response to determining that the message for effecting the new registration indicates a UE registration pattern anomaly, blocking the message for effecting the new registration.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The subject matter described herein will now be explained with reference to the accompanying drawings of which:

(2) FIG. 1 is a network diagram illustrating an exemplary 5G network architecture;

(3) FIG. 2 is a network diagram illustrating an exemplary attack scenario where an attacker sends a fake UE registration to the home network to direct subsequent communications for the UE to the attacker;

(4) FIG. 3 is a network diagram illustrating a network data analytics function (NWDAF) or aggregation node that generates mobility patterns for UEs and an SEPP that screens UE registrations using mobility patterns;

(5) FIG. 4 is a network diagram of the same network illustrated in FIG. 3 where a UE registration is screened based on mobility pattern information and the UE registration is allowed because the SEPP determines from the mobility pattern information that the registration is not anomalous;

(6) FIG. 5 is a network diagram illustrating the same network as FIG. 3 where a UE registration is screened based on mobility pattern information and the UE registration is blocked because the SEPP determines from the mobility pattern information that the registration is anomalous;

(7) FIG. 6 is a block diagram illustrating an exemplary architecture for an aggregation node and an SEPP for performing 5G historical mobility tracking and security screening based on mobility patterns;

(8) FIG. 7 is a flow chart illustrating an exemplary process performed by an aggregation node in generating mobility patterns and providing the mobility patterns to a requesting node;

(9) FIG. 8 is a flow chart illustrating exemplary process performed by an SEPP in obtaining mobility patterns from an aggregation node and screening UE registrations based on the mobility patterns;

(10) FIG. 9 is a signaling message flow diagram illustrating exemplary signaling associated with 5G mobility; and

(11) FIG. 10 is a network diagram illustrating an exemplary unified data repository (UDR) for performing 5G mobility tracking according to an aspect of the subject matter described herein.

DETAILED DESCRIPTION

(12) The subject matter described herein relates to methods, systems, and computer readable media for 5G UE historical mobility tracking and security screening using mobility patterns. FIG. 1 is a block diagram illustrating an exemplary 5G system network architecture. The architecture in FIG. 1 includes NRF 100 and SCP 101, which may be located in the same home public land mobile network (HPLMN). As described above, NRF 100 may maintain profiles of available producer NF service instances and their supported services and allow consumer NFs or SCPs to subscribe to and be notified of the registration of new/updated producer NF service instances. SCP 101 may also support service discovery and selection of producer NF instances. SCP 101 may perform load balancing of connections between consumer and producer NFs. In addition, using the methodologies described herein, SCP 101 may perform preferred NF location based selection and routing.

(13) NRF 100 is a repository for NF or service profiles of producer NF instances. In order to communicate with a producer NF instance, a consumer NF or an SCP must obtain the NF or service profile or the producer NF instance from NRF 100. The NF or service profile is a JavaScript object notation (JSON) data structure defined in Third Generation Partnership Project (3GPP) Technical Specification (TS) 29.510. The NF or service profile definition includes at least one of a fully qualified domain name (FQDN), an Internet protocol (IP) version 4 (IPv4) address or an IP version 6 (IPv6) address. In FIG. 1, any of the nodes (other than NRF 100) can be either consumer NFs or producer NFs, depending on whether they are requesting or providing services. In the illustrated example, the nodes include a policy control function (PCF) 102 that performs policy related operations in a network, a user data management (UDM) function 104 that manages user data, and an application function (AF) 106 that provides application services. The nodes illustrated in FIG. 1 further include a session management function (SMF) 108 that manages sessions between access and mobility management function (AMF) 110 and PCF 102. AMF 110 performs mobility management operations similar to those performed by a mobility management entity (MME) in 4G networks. An authentication server function (AUSF) 112 performs authentication services for user equipment (UEs), such as user equipment (UE) 114, seeking access to the network.

(14) A network slice selection function (NSSF) 116 provides network slicing services for devices seeking to access specific network capabilities and characteristics associated with a network slice. A network exposure function (NEF) 118 provides application programming interfaces (APIs) for application functions seeking to obtain information about Internet of things (IoT) devices and other UEs attached to the network. NEF 118 performs similar functions to the service capability exposure function (SCEF) in 4G networks.

(15) A radio access network (RAN) 120 connects user equipment (UE) 114 to the network via a wireless link. Radio access network 120 may be accessed using a g-Node B (gNB) (not shown in FIG. 1) or other wireless access point. A user plane function (UPF) 122 can support various proxy functionality for user plane services. One example of such proxy functionality is multipath transmission control protocol (MPTCP) proxy functionality. UPF 122 may also support performance measurement functionality, which may be used by UE 114 to obtain network performance measurements. Also illustrated in FIG. 1 is a data network (DN) 124 through which UEs access data network services, such as Internet services.

(16) SEPP 126 filters incoming traffic from another PLMN and performs topology hiding for traffic exiting the home PLMN. SEPP 126 may communicate with an SEPP in a foreign PLMN which manages security for the foreign PLMN. Thus, traffic between NFs in different PLMNs may traverse two SEPP functions, one for the home PLMN and the other for the foreign PLMN.

(17) As stated above, one problem that occurs in 5G networks is that attackers can spoof the identity of a network serving a mobile subscriber in a UE registration message, cause the home network to store a false location for the UE, and, as a result, subsequent communications intended for the real UE can be directed for an attacker. FIG. 2 illustrates an exemplary attack scenario where an attacker sends a fake UE registration and uses the fake UE registration to intercept communications to a UE. In addition to the 5G NFs described above with respect to FIG. 1, in FIG. 2, the network includes. radio access network nodes 202A, 202B, 204A, and 204B. Radio access network nodes 202A, 202B, 204A, and 204B enable wireless communications with UEs. The architecture of radio access network nodes 202A, 202B, 204A, and 204B depends on how the gNB function in the network is implemented. gNB functionality may be implemented using a base band unit (BBU) where the functionality of the gNB is not split into separate nodes or split into separate distributed unit (DU) and control unit (CU) nodes. The CU is a logical node that includes gNB functions like transfer of user data, mobility, control, radio access network sharing positioning, etc. The DU is the logical node that implements a subset of the gNB functions depending on the functional split. The operation of the DU is controlled by the CU. The radio unit (RU) is just another name for the CU.

(18) Referring to the message flow in FIG. 2, in step 1, an attacker 200 impersonates an AMF by sending a fake UE registration to the home network of a UE. The fake UE registration identifies the UE location as the attacker's network. However, the UE is not present in the attacker's network.

(19) An SEPP 126A located in the UE's home network receives the fake UE registration. In this example, it is assumed that SEPP 126A does not implement security screening based on mobility pattern information. Accordingly, in step 2, SEPP 126A in the home network sends the fake UE registration to UDM 104 in the home network.

(20) UDM 104 likewise does not implement security screening of UE registrations based on mobility patterns. Accordingly, in step 3, UDM 104 updates the location of the UE in the database record maintained by the UDM for the UE to be the attacker network. The location of the UE stored by UDM 104 is used by other nodes to communicate with the UE. However, because the location stored for the UE in FIG. 2 is fake, the registration of the fake location will cause the communications will be redirected to the attacker.

(21) In step 4, a legitimate caller UE 206 initiates communications (e.g., a voice call or text message) with the UE whose registration has been compromised. The initiation of communications causes AMF 110B located in the calling UE's network to send a location query to the home network of the UE (step 4). SEPP 126B located in the caller's network forwards the location query to the called UE's home network. SEPP 126A receives the location query and, in step 5, forwards the location query to UDM 104.

(22) In step 6, UDM 104 performs a lookup in its database using the UE identified in the location query and generates a location response including the registered location of the UE. In this case, the registered location is the attacker's network. UDM 104 sends the response with the fake UE location to the requesting AMF via SEPP 126A and SEPP 126B. SEPP 126B forwards the response with the fake UE location to AMF 110B. In step 7, AMF 110B forwards communications from UE 206 to the location specified in the location response message, which, in this case, is the location of attacker 200. Thus, without some security screening, attacker 200 can redirect voice calls, text messages, or other communications intended for a UE whose registration has been compromised to the attacker's network.

(23) To reduce the likelihood of successful redirection of communications to an attacker's network, the subject matter described herein includes collecting mobility pattern information for a UE and using the mobility pattern information to perform security screening of registrations sent for a UE. FIG. 3 is a network diagram illustrating the use of an aggregation node to generate mobility pattern information that is usable to screen for fake UE registrations. Referring to FIG. 3 an aggregation node 300 may be a computing platform that collects UE registration information from registrations of UEs 302, 304, and 306 as the UEs register with different networks. In the illustrated example, the UEs include IoT devices and mobile computing devices, including mobile handsets, tablets, laptops, etc. Aggregation node 300 may be implemented using a computing platform capable of collecting UE registration data, aggregating the registration data, producing mobility patterns from the UE registration data, and providing the mobility patterns to querying NFs for UE registration security screening. In one example, aggregation node 300 may be implemented using an NWDAF as defined in 3GPP TS 29.520. The NWDAF allows consumer NFs to subscribe to receive notification of events from the NWDAF and also to request for one time notification of an event. The request-response communications model is more applicable to the architecture in FIG. 3 where the SEPP requests a mobility pattern from aggregation node 300 in response to receiving a new UE registration. It should also be noted that although aggregation node 300 could can be implemented by an NWDAF, the subject matter described herein is not limited to using an NWDAF to generate mobility patterns. In an alternate implementation, an non-3GPP defined network function may be used to receive the UE registration data, generate the mobility patterns, and provide the mobility patterns to requesting network functions.

(24) In the message flow illustrated in FIG. 3, when any of UEs 302, 304, or 306 connects to a network via one of radio access network nodes 202A and 202B or via Wi-Fi/broadband/satellite gateway 308, the serving AMF sends a UE registration to UDM 104 located in the home network of the UE and sends a copy of the registration or registration data extracted from the UE registration to network data aggregation node 300. In FIG. 3, in step 1, UEs 304 and 306 connect to the home network, and home AMF 110A sends a copy of the UE registration information for each registration to network data aggregation node 300. The UE registration information includes an indicator of the UE location, which in this example is the home network of the UE, the timestamp of the registration, and the type allocation code (TAC), which identifies the UE device type. In the illustrated example, the registration information for UE 304 includes the UE location identifier of the home network, a timestamp of TS1, and a TAC of 1. The registration information for UE 306 includes a UE location identifier that identifies the home network, a timestamp of TS1.1, and a TAC of 2, since UE 306 is an IoT device, which is a different type of device than UE 304.

(25) In step 2, network data aggregation node 300 receives or updates mobility pattern information for the registering UEs. For example, network data aggregation node 300 may maintain a mobility pattern database for all UEs that network data aggregation node 300 tracks. When network data aggregation node 300 receives registration data regarding a UE, network data aggregation node 300 performs a lookup in its mobility pattern database to determine if a mobility pattern record exists for the UE. If a record does not exist, network data aggregation node 300 may generate a new mobility pattern record for the UE. If a record exists, network data aggregation node 300 may update the existing mobility pattern record for the UE based on the newly received UE registration data.

(26) In the illustrated example, it is assumed that the registrations are new (first time) registration for UEs 304 and 306. Accordingly, network data aggregation node 300 will create the following mobility pattern records for UEs 304 and 306:

(27) TABLE-US-00001 TABLE 1 Mobility Pattern for UE 304 after First Registration Location UE ID Location Timestamp TAC UE1 Home NW TS1 1

(28) TABLE-US-00002 TABLE 2 Mobility Pattern for UE 306 after First Registration Location UE ID Location Timestamp TAC UE2 Home NW TS1.1 2

(29) Continuing with the example in FIG. 3, UEs 304 and 306 roam into a visited network and connect to the visited network via radio access network nodes 204B and 202B. In step 3, the serving AMF 110B sends new UE registrations (one for each of UEs 304 and 306) to the home network of UEs 304 and 306 and sends a copy of the UE registration data to network data aggregation node 300. In the illustrated example, the aggregation data includes an identifier of the visited network as the UE location, a timestamp of TS2, and a TAC of 1 and 2, respectively, for UEs 304 and 306.

(30) Network data aggregation node 300 receives the UE registration data, determines that mobility pattern records already exist for the UEs, and i[dates the mobility pattern records for UEs 304 and 306. Tables 3 and 4 below illustrate the updated mobility patterns that may be stored for UEs 304 and 306.

(31) TABLE-US-00003 TABLE 3 Updated Mobility Pattern for UE 304 after Second Registration Location UE ID Location Timestamp TAC UE1 Home NW TS1 1 Visited NW1 TS2 1

(32) TABLE-US-00004 TABLE 4 Updated Mobility Pattern for UE 306 after Second Registration Location UE ID Location Timestamp TAC UE2 Home NW TS1.1 2 Visited NW1 TS2.1 2

(33) The data illustrated in Tables 1-4 may be used by requesting network nodes to determine whether a registration for a UE is anomalous and to perform security screening for UE registrations. FIG. 4 illustrates the use of the mobility pattern to screen a UE registration where the UE registration is determined not to be anomalous and is allowed. Referring to FIG. 4, in step 1, it is assumed that UE 304 roams from the home network into a visited network. When UE 304 connects to the visited network via radio access nodes 204B and 202B, serving AMF 110B sends a UE registration message to UDM 104 located in the home network of UE 304. The UE registration identifies the UE location as visited network #1 and includes the timestamp TS3 and the TAC of 1. In step 2, the UE registration is forwarded to the home network via SEPP 126B in the visited network and SEPP 126A in the UE's home network.

(34) In step 3, SEPP 126A in the home network, in response to receiving the UE registration, sends a mobility pattern query to aggregation node 300. Aggregation node 300 responds to the mobility pattern query with a query response in step 4 that includes the mobility pattern.

(35) SEPP 126A compares the UE location, timestamp, and type allocation code in the registration to the mobility pattern received from aggregation node 300. In this example, it is assumed that the UE is visiting a network that the UE has visited before. Accordingly, SEPP 126A determines that the UE registration is not anomalous and, in step 5, forwards the UE registration to UDM 104 in the UE's home network.

(36) FIG. 5 illustrates the case where a UE registration is blocked based on mobility pattern information. Referring to FIG. 5, in step 1, an attacker, 200 emulates an AMF by sending a fake UE registration to the home network to home network of a UE. In the fake UE registration, the attacker includes an identifier of the of the attacker's network instead of the location where a real UE is registered.

(37) In step 2, SEPP 126A receives the fake UE registration and sends a mobility pattern query to aggregation node 300. The mobility pattern query identifies the UE and the location of the UE contained in the registration message.

(38) Aggregation node 300 receives the mobility pattern query, performs a lookup in its mobility pattern database using the UE identifier obtained from the mobility pattern query, and retrieves a mobility pattern for the UE. In step 3, aggregation node 300 responds to the mobility pattern query with a query response that includes the mobility pattern. The response may optionally include an indication that the UE registration is anomalous based on a determination made by aggregation node 300.

(39) SEPP 126A receives the response and determines that the response is anomalous. In a simple example, SEPP 126A may determine that the response is anomalous if the UE location specified in the registration does not match any of the previous UE registration locations indicated in the mobility pattern. In this example, SEPP 126A determines that the mobility pattern does not correspond to a network where the UE has been registered in the past. Accordingly in step 4, SEPP 126A blocks the fake UE registration. SEPP 126A may also implement further verification before blocking the UE registration. Examples of further verification will be described in more detail below.

(40) FIG. 6 is a block diagram illustrating an exemplary architecture for aggregation node 300 and SEPP 126A for performing security screening using mobility patterns. Referring to FIG. 6, aggregation node 300 and SEPP 126A each include at least one processor 600 and memory 602. Aggregation node 300 includes a mobility pattern generator 604 that generates mobility patterns based on UE registration data received from AMFs and stores the mobility patterns in a mobility pattern database 606. Alternatively, rather than receiving the UE registration data from AMFs in the networks where UEs roam, mobility pattern generator 604 may receive the UE registration data from SEPPs in the network where UEs roam. In yet another alternative, mobility pattern generator 604 may receive the UE registration data used to populate mobility pattern database 606 from home network SEPPs in response to incoming registration requests for roaming subscribers of each SEPP's home network. Mobility pattern generator 604 may utilize the mobility pattern information in mobility pattern database 606 to respond to mobility pattern queries from home network SEPPs.

(41) SEPP 126A includes a UE registration security screener 608 that receives UE registrations from AMFs, formulates queries to network data aggregation node 300, receives mobility pattern information from aggregation node 300, and determines whether to block UE registrations based on the mobility pattern information. UE registration security screener 608 may block UE registrations that are identified as anomalous based on the mobility pattern information. UE registration security screener 608 may also perform UE location verification after an initial determination that a UE registration is anomalous. In one example, UE registration security screener 608 may initiate paging of the UE at the location specified in the UE registration. If the UE is successfully paged, this indicates the presence of a real UE at the location specified in the UE registration, and UE registration security screener 608 may allow the registration. If the UE is not successful paged, UE registration security screener 608 may block the registration as fake. In another example, UE registration security screener 608 may allow registrations identified as anomalous but send a message to the network operator including registration data (network identifier, UE identifier, TAC, etc.) for the anomalous registrations.

(42) FIG. 7 is a flow chart illustrating an exemplary process performed by, aggregation node 300 in generating mobility patterns and providing the mobility patterns to requesting NFs. Referring to FIG. 7, in step 700, aggregation node 300 receives UE registration data from 5G NFs as UEs connect to different network locations. For example, aggregation, node 300 may receive UE registration data, such as registration locations, timestamps, and type allocation codes from AMFs as UEs register with networks served by the AMFs. In addition, or, alternatively, aggregation node 300 may receive mobility patterns as defined in section. 5.3.4.2 of 3GPP TS 23.501. 3GPP TS 23.501 indicates that the mobility pattern is a concept that may be used by the AMF to characterize and optimize UE mobility. The AMF determines and updates the mobility pattern of the UE based on the subscription of the UE, statistics of UE mobility, network local policy, UE assisted information, or any combination of these items. The statistics of UE mobility can be historical or expected UE moving trajectory. If the NWDAF is deployed, the statistics of can also be analytics, i.e., statistics or predictions provided by the NWDAF. According to the subject matter described herein, the mobility pattern information can be used by the SEPP to perform UE registration security screening, which is not mentioned as a use of the mobility pattern in 3GPP TS 23.501.

(43) Returning to FIG. 7, in step 702, the aggregation node aggregates registration data for the UEs received from the 5G NFs to produce mobility patterns for individual UEs. For example, aggregation node 300 may aggregate UE registration data from different AMFs to produce mobility patterns for the UEs, examples of which are illustrated above and Tables 1-4. Aggregation node 300 may store the mobility patterns in mobility pattern database 606.

(44) In step 704, the network data aggregation node receives a mobility pattern query from a 5G NF located in the home network of the UE. For example, SEPP 126A may send a mobility pattern query to aggregation node 300. The mobility pattern query may be generated in response to a request generated by an AMF currently serving the UE to register a UE's current location with the home network UDM or by an attacker impersonating an AMF seeking to update the UE's location to point to the attacker's network. SEPP 126A may cache the UE registration request pending a determination as to whether the registration request is anomalous based on the mobility pattern information. The mobility pattern query may identify the UE for which registration is being requested. The mobility pattern query may optionally include the network location of the UE extracted from the registration request for the case where aggregation node 300 makes a determination as to whether the registration request is anomalous.

(45) In step 706, aggregation node 300 responds to the mobility pattern query by transmitting the mobility pattern to the 5G NF located in the home network of the UE. For example, aggregation node 300 may locate the mobility pattern for the UE in mobility pattern database 606 and transmit the mobility pattern to SEPP 126A located in the home network of the UE. In addition, aggregation node 300 may provide an indication in the response and to whether aggregation node 300 classifies the registration as anomalous.

(46) FIG. 8 is a flow chart illustrating an exemplary process performed by a home network SEPP and performing security screening based on mobility pattern information. Referring to FIG. 8, in step 800, the SEPP receives a request to register a UE. For example, SEPP 126A may receive a registration request for a legitimate UE or a fake registration from an attacker.

(47) In step 802, in response to the registration request, the SEPP transmits a mobility pattern query message to the network data aggregation node. For example, SEPP 126A may transmit a mobility pattern query message to aggregation node 300. The mobility pattern query message may identify the UE whose registration is being requested and optionally the location for which the registration is requested. The location may be provided for the case where aggregation node 300 makes a determination as to whether the registration is anomalous. If the determination is made solely by SEPP 126A, then sending the location of the UE as part of the mobility pattern query is not required. SEPP 126A may cache and refrain from forwarding the registration request to the UDM in the home network while the determination as to whether the registration request is anomalous is being made.

(48) In step 804, the SEPP 126A receives, from the network data aggregation node, a response, including the mobility pattern for the UE. For example, SEPP 126A may receive mobility pattern information, such as that illustrated above in Tables 1-4, from aggregation node 300. The mobility pattern information may indicate a historical pattern of registrations for the UE identified in the request. The mobility pattern may additionally or alternatively indicate a mobility pattern for UEs of the same device type as the UE identified in the mobility pattern request. Mobility patterns for UEs of the same device type may be useful in determining whether a registration for a particular UE, such as an IoT device, is anomalous, when compared with mobility patterns of other UEs of the same device type. For example, one would expect a registration for a mobile handset with voice call capabilities to have different ranges of values for various parameters in the registration request than a registration from an IoT device where the IoT device is a sensor. Another check for an IoT device could be to determine whether the device type is mobile. For example, a stationary IoT device may be expected to always register from the same location when the device wakes up to send its data. A registration for the same stationary IoT device for network locations may indicate that the registration is anomalous.

(49) In step 806, SEPP 126A determines, based on the mobility pattern, whether the registration is anomalous. For example, SEPP 126A may identify the registration as being anomalous if the location of the registration indicates a location in a network where the UE has never been registered in the past. In another example, SEPP 126A may implement a more sophisticated algorithm, such as a machine learning algorithm, to determine whether a UE registration is anomalous.

(50) In step 808, if the registration is not determined to be anomalous, control proceeds to step 810 where the registration is allowed. If the registration is allowed, SEPP 126A will forward the registration request to the UDM in the home network of the UE. The UDM will update the UE's location in its database, and subsequent communications to the UE will be sent to that location until the registration is canceled or updated by a new registration.

(51) If the registration is determined to be anomalous, as indicated above, in some instances, the SEPP may implement further verification to determine whether to allow the registration. For example, if the UE is a mobile telephone that is registering for the first time in a new city, the registration may be valid if the user has traveled to that city for the first time. Accordingly, in step 812, it is determined whether further verification of the registration is implemented. If further verification is not implemented, control proceeds to step 814 where the SEPP blocks the registration. Blocking the registration may include refraining from forwarding the registration to the UDM, and, optionally responding to the requesting node (real AMF or attacker).

(52) In step 812, if further verification is implemented control proceeds to step 816 where further verification of the UE's registration is performed. In one example, SEPP 126A may initiate paging of the UE. Initiating paging of the UE may include sending a paging request message to the UE via the network location identified in the registration request. The paging response will be sent back to the requesting SEPP 126A. If the paging response indicates that the UE was successfully paged, this is an indication that there is a real UE at the location specified in the registration request.

(53) Accordingly, in step 818, if the UE location is verified, control proceeds to step 820 where the registration is allowed. If the SEPP 126A receives a response indicating that the paging was unsuccessful, this is an indication that there is not a real UE at the location specified in the registration request and that the registration is fake. Accordingly, if the UE location is not verified by receiving a response indicating successful paging of the UE, control proceeds to step. 814 where the registration is blocked.

(54) It should be noted that subject matter described herein is not limited to blocking UE registrations that are identified as anomalous. In addition to or instead of blocking the registrations, SEPP 126A may notify the network operator by transmitting a message to an operations, administration, and maintenance (OA&M) system of the network operator. In the case where a registration is identified as anomalous but is allowed, the network operator may choose to block subsequent registrations that identify the same UE if it is determined after further analysis that the registrations are fake.

(55) The subject matter described herein is not limited to identifying a UE registration as anomalous based on a comparison of the historical mobility pattern for the same UE. In an alternate implementation, network data aggregation node 300 may collect UE registration data for UEs of the same type and provide, to a requesting node, the UE registration data for the UEs of the same type as a UE attempting to register. In such a case, the requesting node, such as home network SEPP 126A will compare the registration data of the UE attempting to register with UE mobility pattern(s) of UEs of the same type as the UE requesting to register. If the registration data of the UE attempting to register is statistically different from the mobility patterns of the UEs of the same type, the UE registration may be blocked. If the registration data of the UE attempting to register is not statistically similar to the mobility patterns of the UEs of the same type, the registration may be blocked.

(56) As stated above, the network data aggregation node 300 is configured to perform 5G historical mobility tracking using signaling messages exchanged between 5G network nodes to support 5G UE mobility. FIG. 9 is a signaling message flow diagram illustrating exemplary 5G messages associated with mobility of a UE. Referring to FIG. 9, when UE 304 connects to the network via RAN 120, RAN 120 performs AMF selection and sends a registration request message to serving AMF 110. Serving AMF 110B performs AUSF selection and signals with AUSF 112 in the home network of the UE to authenticate the UE. Once the UE is authenticated, serving AMF sends an Nudm_UECM_Registration_Request message 900 to UDM 104 located in the home network of UE 304. The message may be sent via an SEPP (not shown in FIG. 9) in the network of AMF 110B. Serving AMF 110B may send a copy of the registration data from Nudm_UECM_Registration_Request message 900 to network data aggregation node 300, and network data aggregation node 900 may store the registration data in UE mobility pattern database 606. The home network SEPP (not shown in FIG. 9) may query the UE mobility pattern database to determine whether to allow Nudm_UECM_Registration_Request message 900 to be sent to UDM 104 to register the user. The remaining signaling in FIG. 9 relates to policy control and session setup for the UE once the UE's registration is allowed to proceed with the UDM. If the registration is blocked, the remaining signaling after the registration is avoided.

(57) In the examples above, aggregation node 300 is described as being implemented using an NWDAF or a non-3GPP defined aggregation platform. In another example, aggregation node 300 can be implemented using a 5G UDR. FIG. 10 is a network diagram illustrating a 5G UDR and its interfaces with other 5G network nodes. In FIG. 10, UDR 1000 stores subscription data, policy data, structured data for exposure, and application data. This data can be collected from UDM 104, PCF 102, and NEF 118. The interfaces between UDM 104 and AMF 110, SMF 108, AUSF 112, and short message service function (SMSF) 1002 are also illustrated.

(58) If UDR 1000 implements the functionality of aggregation node 300 described above, UDR 1000 may utilize UE registration data obtained from UDM 104 to populate mobility pattern database 606. In addition, UDR 1000 may utilize the data in mobility pattern database 606 to respond to mobility pattern queries, as described above.

(59) Thus, using historical UE registration information to generate mobility patterns and using the mobility patterns to screen UE registrations, some network attacks can be avoided and data integrity in the home network is enhanced. Using mobility patterns is more computationally efficient than computing distances between networks and using the distances to perform velocity checks.

(60) Implementing the mobility pattern generation at a network data aggregation platform, such as an NWDAF, UDR, or non-3GPP data aggregation platform, is advantageous because such an implementation offloads the storage, processing and retrieval required to generate the mobility patterns from nodes, such as AMFs, that are also responsible for establishing and maintaining sessions between UEs. Implementing security screening at the home network SEPP is advantageous because the SEPP is located at a natural ingress and egress point of the network.

(61) Even though in the examples described above the mobility patterns are generated by the network data aggregation node and the determination as to whether a registration is anomalous is made by the SEPP, the subject matter described herein is not limited to such an implementation. In an alternate implementation, the mobility pattern generation and determination as to whether a registration is anomalous may be performed by the network data aggregation node. In such an implementation, the message flow may be the same as that illustrated in FIG. 3 except the response from the network data aggregation node to the mobility pattern query may include a determination as to whether the registration is anomalous or not. Upon receiving the response with the determination, the SEPP may 1) accept the determination, 2) perform UE location verification as described above and accept or overturn the determination based on results of the verification. That is, if the determination made by the network data aggregation node is that a registration is anomalous, but the UE's location is verified by a successful paging of the UE, the SEPP may record the determination made by the network data aggregation node, record that the UE was successfully paged, and allow the registration to proceed into the home network of the UE. If the determination made by the network data aggregation node is that a registration is anomalous, and the UE is not successfully paged, the SEPP may block the registration. If the determination by the network data aggregation function is that the registration is not anomalous, the SEPP may allow the registration or only allow the registration after successful verification of the UE location though successful paging.

(62) The disclosure of each of the following references is hereby incorporated herein by reference in its entirety:

REFERENCES

(63) 1. 3GPP TS 23.501; 3.sup.rd Generation Partnership Project; Technical Specification Group Services and System Aspects; System architecture for the 5G System (5GS); Stage 2 (Release 16) V16.4.0 (2020 March) 2. 3GPP TS 23.502; 3.sup.rd Generation Partnership Project; Technical Specification Group Services and System Aspects; Procedures for the 5G System (5GS); Stage 2 (Release 16) (2020 March) V16.4.0 (2020 March) 3. 3GPP TS 29.510; 3.sup.rd Generation Partnership Project; Technical Specification Group Core Network and Terminals; 5G System; Network Function Repository Services; Stage 3 (Release 16), V16.4.0 (2020 July) 4. 3GPP TS 29.500; 3.sup.rd Generation Partnership Project; Technical Specification Group Core Network and Terminals; 5G System; Technical Realization of Service Based Architecture; Stage 3 (Release 16), V16.4.0 (2020 June) 5. 3GPP TS 29.520, 3.sup.rd Generation Partnership Project, Technical Specification Group Core Network and Terminals; 5G System; Network Data Analytics Services; Stage 3 (Release 16), V16.4.0 (2020 June)

(64) It will be understood that various details of the presently disclosed subject matter may be changed without departing from the scope of the presently disclosed subject matter. Furthermore, the foregoing description is for the purpose of illustration only, and not for the purpose of limitation.