hash stringlengths 32 32 | doc_id stringlengths 5 12 | section stringlengths 5 1.47k | content stringlengths 0 6.67M |
|---|---|---|---|
ca180014f5532ca12c04dc24c74a1668 | 33.714 | 5.6.3 Evaluation
| Editor’s Note: Evaluation is FFS
|
ca180014f5532ca12c04dc24c74a1668 | 33.714 | 5.7 DO-A Capable AIoT device identifier protection with Bloom filter
| |
ca180014f5532ca12c04dc24c74a1668 | 33.714 | 5.7.1 Introduction
| The solution addresses Key Issue #4: AIOT device ID protection in DO-A procedure. The solution introduces a Bloom filter-based procedure, together with filtering information, to efficiently page multiple DO-A Capable AIoT devices while protecting identifier privacy.
In this solution, the ADM determines the actually paged target devices from the Tag, rather than processing every device included in the filtering information. As a result, the ADM only operates on the devices involved in the current paging round, which reduces unnecessary matching and lookup steps and lowers the overall processing overhead on the ADM side.
In addition, the proposed solution does not require the transmission of AIoT device permanent identifiers, thereby enhancing identifier privacy, enables the simultaneous paging of multiple AIoT devices in a single procedure, thereby reducing signalling load on the air interface and between network entities, supports the indication of multiple, non-related AIoT devices within a single fixed-length Tag, and prevents non-target devices from responding.
|
ca180014f5532ca12c04dc24c74a1668 | 33.714 | 5.7.2 Solution details
| |
ca180014f5532ca12c04dc24c74a1668 | 33.714 | 5.7.2.1 Procedure
| The purpose of this solution is to enable multiple inventory operations while protecting the AIoT device permanent identifier during the AIoT device inventory procedure.
Figure 5.7.2.1-1: Inventory procedure
0. Step 1-6 of clause 6.2.2 Procedure for Inventory or clause 6.2.3 Procedure for command in TS 23.369 [7] is performed.
1.The AIOTF sends the filtering information to the ADM.
2 The ADM generates the TagN for indicating the target devices to be paged [see 5.7.2.1].
Editor’s Note: It is FFS that if and how freshness is added for TagN generation.
3. The ADM sends the Tag to the AIOTF.
4.The AIOTF sends an inventory request message including the Tag and the filtering information to the NG-RAN.
5. NG-RAN includes the Tag and the filtering information in the paging request message to the AIoT device.
6. Upon receiving the paging request message, the AIoT device checks whether it is within the device range indicated by the filtering information. If it is within the range, the device retrieves its bit indices. It considers itself paged only if the bit values at those indices in the Tag are all “1”, otherwise, it should ignore the paging message.
The AIoT device shall pre-store the Tag bit indices. In most use cases, 3 to 10 indices are sufficient. In practice, the optimal value can be calculated as k = (m / n) ln 2, where m is the Tag length, n is the number of target devices.
7. AIoT device sends D2R message to the NG-RAN.
8. NG-RAN sends Inventory report message to AIOTF.
9-10. [Optional] If the received AIoT Identification Information needs to be decrypted by ADM, the AIoTF sends the Tag and the filtering information to the ADM to fetch the device permanent identifier.
Editor’s Note: Whether a Bloom filter can be constructed that can be accommodated within a constrained group paging message while maintaining acceptable false positive rate is FFS.
Editor’s Note: The alignment with SA2 procedure (e.g. inventory, command) is FFS.
|
ca180014f5532ca12c04dc24c74a1668 | 33.714 | 5.7.2.2 Tag generation
| 1. TagD generation
The following parameters shall be used to form the input S to the k KDFs:
- FC = 0xNN,
- P0 = Device permanent identifier,
- L0 = length of Device permanent identifier,
The input key KEY shall be KAIOT_root. The P0 input is the stored AIoT device permanent identifier. The outputs of the k KDFs are denoted T1, T2, …, Tk.
The TagD of each device is an m-bit array (index range 0…m−1) that is initialized to all zeros. Each output Tj (for j = 1…k) shall be mapped to a bit index idxj as idxj = Tj mod m. The bit in TagD at index idxj shall be set to 1.
2. TagN constructed by the ADM
For all n target devices (i.e., device 1 to device n) in an inventory round, the ADM constructs TagN, its bit array is formed by performing a bitwise OR (union) on the set bits from the TagD of each device.
3. The indices of TagD pre-stored in the device
Each AIoT device has pre-stored the bit indices of its own TagD, so that upon receiving a paging message carrying the TagN, it can directly check those positions in the TagN to determine whether it is being paged.
|
ca180014f5532ca12c04dc24c74a1668 | 33.714 | 5.7.3 Evaluation
| TBD
|
ca180014f5532ca12c04dc24c74a1668 | 33.714 | 5.8 Solution #8: SUCI
| |
ca180014f5532ca12c04dc24c74a1668 | 33.714 | 5.8.1 Introduction
| This solution addresses Key Issue #4 and applies to topology 1 and topology 2.
|
ca180014f5532ca12c04dc24c74a1668 | 33.714 | 5.8.2 Solution details
| This solution proposes the use of SUCI (Subscription Concealed Identifier), as specified in TS 33.501 [9], to protect the AIoT device permanent ID. The SUCI is calculated with non-null scheme.
Editor’s note: how to protect AIoT device permanent ID in SNPN is FFS
Editor’s note: Whether AIoT devices have capability to perform SUCI calculation is FFS
|
ca180014f5532ca12c04dc24c74a1668 | 33.714 | 5.8.3 Evaluation
| This solution addresses Key Issue #4 thanks to ID protection mechanism already specified in TS 33.501 [9].
The possibility to perform SUCI calculation depends on AIoT device capability.
Editor’s Note: Further evaluation is FFS.
5.Y Solution #Y: <Solution Name>
5.Y.1 Introduction
Editor’s Note: Each solution should list the key issues being addressed.
5.Y.2 Solution details
5.Y.3 Evaluation
Editor’s Note: Each solution should motivate how the potential security requirements of the key issues being addressed are fulfilled.
|
ca180014f5532ca12c04dc24c74a1668 | 33.714 | 6 Conclusions
| Editor’s Note: This clause captures the conclusions of this study.
Annex <X>:
Change history
Change history
Date
Meeting
TDoc
CR
Rev
Cat
Subject/Comment
New version
10/2025
SA3#124
S3‑253300
Initial draft TR
0.0.1
10/2025
SA3#124
S3‑253732
Incorporated accepted contributions S3‑253822, S3‑253823, S3-253824, S3-253825, S3-253826, S3-253827
0.1.0
11/2025
SA3#125
S3‑254541
Incorporated accepted contributions S3-254694, S3-254695, S3-254696, S3-254697, S3-254698, S3-254699, S3-254700, S3-254701, S3-254702, S3-254703, S3-254705, S3-254706
0.2.0
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 1 Scope
| The present document studies the security architecture and security requirements for WAB-nodes, security impacts of potentially compromised WAB nodes and requirements for countermeasures against any compromised WAB nodes.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 2 References
| The following documents contain provisions which, through reference in this text, constitute provisions of the present document.
- References are either specific (identified by date of publication, edition number, version number, etc.) or non‑specific.
- For a specific reference, subsequent revisions do not apply.
- For a non-specific reference, the latest version applies. In the case of a reference to a 3GPP document (including a GSM document), a non-specific reference implicitly refers to the latest version of that document in the same Release as the present document.
[1] 3GPP TR 21.905: "Vocabulary for 3GPP Specifications".
[2] 3GPP TS 23.501: "System architecture for the 5G System (5GS)".
[3] 3GPP TS 38.401: "NG-RAN Architecture description".
[4] 3GPP TS 33.501: "Security architecture and procedures for 5G System".
[5] 3GPP TR 33.745: "Study on security aspects of 5G Next Radio (NR) Femto".
[6] 3GPP TS 33.320: "Security of Home Node B (HNB) / Home evolved Node B (HeNB)".
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 3 Definitions of terms, symbols and abbreviations
| |
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 3.1 Terms
| For the purposes of the present document, the terms given in 3GPP TR 21.905 [1] and the following apply. A term defined in the present document takes precedence over the definition of the same term, if any, in 3GPP TR 21.905 [1].
example: text used to clarify abstract rules by applying them literally.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 3.2 Symbols
| For the purposes of the present document, the following symbols apply:
<symbol> <Explanation>
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 3.3 Abbreviations
| For the purposes of the present document, the abbreviations given in 3GPP TR 21.905 [1] and the following apply. An abbreviation defined in the present document takes precedence over the definition of the same abbreviation, if any, in 3GPP TR 21.905 [1].
<ABBREVIATION> <Expansion>
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 4 Security Architecture and Assumptions
| Editor’s Note: This clause contains security architecture and assumptions to be considered for the study (e.g., per work task/KI).
Figure 5.49.1.1-1 in TS 23.501[2] shows the MWAB architecture for 5GS. In the architecture. There are two components in MWAB, i.e. MWAB-gNB and MWAB-UE. The WAB-node integration procedure is captured in TS 38.401[3].
From a security point of view, the MWAB architecture rely on the 5G security framework for key management and authorization as captured in TS 33.501 [4].
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5 Key issues
| Editor’s Note: This clause contains all the key issues identified during the study.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.1 Key Issue #1: Security of the link between WAB-gNB and OAM
| |
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.1.1 Key issue details
| Based on the WAB-node integration procedure, the WAB-gNB will receive the OAM of WAB through the WAB-MT’s network. The link between WAB-gNB and OAM needs to have sufficient security protection for configuration data transmission.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.1.2 Security threats
| If the link between WAG-gNB and OAM is not well protected, the configuration data will be tampered or disclosure.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.1.1 Potential security requirements
| The link between the MWAB-gNB and the OAM shall be ciphering and integrity protected.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.2 Key Issue #2: Security Protection of Compromised WAB Nodes and Core Network Measures
| |
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.2.1 Key issue details
| Wireless Access Backhaul (WAB) nodes, consist of a WAB-gNB (gNB-like functionality) and a WAB-MT (UE-like functionality). These nodes operate in non-trusted environments and may serve as moving backhaul nodes for the 5GS, establishing NG, Xn, and OAM interfaces over PDU sessions through 3GPP backhauls. While 3GPP TR 33.745 [5] studied NR Femto security and reused procedures from TS 33.320 [6], security concerns specific to WAB nodes particularly compromised WAB nodes in untrusted environments remain unaddressed.
Additionally, core network components may not be equipped to detect anomalous behavior from compromised WAB-gNBs, due to the decentralized and mobile nature of such nodes. The compromised WAB nodes could lead to topology poisoning, signalling storms, or user-plane hijacking.
This key issue aims to address the security issues introduced by compromised WAB nodes, where failure to protect the integrity, authenticity, and confidentiality of messages delivered from WAB-gNB and WAB-MT components can expose the 5GS to topology spoofing, rogue signalling, and persistent infiltration.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.2.2 Security threats
| Potential security threat:
• Rogue WAB-gNB Injection: A compromised WAB node may inject unauthorized signalling or reroute traffic maliciously, particularly via spoofed message. Furthermore, a compromised WAB-gNB can attempt to broadcast unauthorized network identifiers or initiate rogue Xn association attempts with neighbouring gNBs causing service disruption.
• Topology Manipulation and Signalling Abuse: Moving WAB nodes may falsely report neighbour relationships via Xn or behave inconsistently across locations, leading to incorrect handover decisions, topology poisoning, or signalling loops.
• Persistent Threat via Dual Roles: Since WAB-MT behaves like a UE and WAB-gNB like a gNB, a compromised WAB can act in both roles to stage cross-layer attacks, bridging between RAN and CN trust domains.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.2.3 Potential security requirements
| The 3GPP system shall support security mechanisms to mitigate risks from compromised WAB nodes, preventing topology spoofing, rogue signalling, and mobility-related traceability threats.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.3 Key Issue #3: Ensuring secure N2, N3 and Xn interfaces for MWAB nodes
| |
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.3.1 Key issue details
| According to the architecture in 23.501[2],the MWAB-gNB establishes the N2 interface with UE’s 5GC, and setup a Xn link with a traditional gNB.
Figure 5.3-1: Architecture for MWAB operation support - non-roaming with one PLMN
A MWAB may be mounted on a moving vehicle and may serve UEs inside or outside the vehicle. A MWAB cannot provide a connection service to a UE unless it establishes secure N2, N3 and Xn connections with UE’s AMF and UE’ NG-RAN.
Since the NDS/IP is used for Xn and N2 connection as defined in TS 33.501[4], credentials must be provided in MWAB case in order to build the connection with UE’s network. One existing method is the preconfigure the potential serving UE’s credentials to the MWAB before it starts to connecting to the network. However, this relies on a well and unchanged plan on the MWAB. Since it is very difficult to know where the MWAB will go, it is very difficult to a MWAB vendor to configure everything in advance. Moreover, moving MWAB nodes can also lead to un-reliable transport between the MWAB node and backhaul. It is important to ensure security while allowing mobility of MWAB nodes.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.3.2 Security threats
| Lack of end-to-end protection for MWAB-gNB’s N2, N3 and Xn can lead to potential tampering of UE related signaling messages and potential breach of confidentiality, integrity and possible availability risks.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.3.3 Potential security requirements
| Credentials for NDS/IP for Xn, N3 and N2 connection between MWAB and UE’s network shall be provided with confidentiality protection and integrity protection.
During movement of MWAB nodes, the end-to-end security of N2, N3 and Xn interfaces shall be ensured.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.4 Key Issue #4: Protection and binding of MWAB-gNB control plane over BH-PDU sessions
| |
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.4.1 Key issue details
| In MWAB, OAM, N2, Xn and N3 traffic for the MWAB-gNB is carried over backhaul PDU session(s) that the MWAB-UE establishes and modifies based on traffic descriptors and OAM configuration. The MWAB broadcasted PLMN/SNPN may differ from the BH PLMN/SNPN which creates inter-PLMN/SNPN trust boundaries for these control plane and OAM links.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.4.2 Security threats
| Interception or modification of OAM/N2/Xn control traffic over BH PDU session(s); misclassification of traffic due to descriptor or mapping error; cross-slice leakage; replay during mobility or BH PDU session changes are possible.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 5.4.3 Potential security requirements
| Confidentiality and integrity protection for OAM/N2/Xn control traffic over BH PDU session(s), binding MWAB-gNB identity and traffic classes to BH PDU sessions.
5.X Key Issue #X: <Key Issue Name>
5.X.1 Key issue details
5.X.2 Security threats
5.X.3 Potential security requirements
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 6 Solutions
| Editor’s Note: This clause contains the proposed solutions addressing the identified key issues.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 6.0 Mapping of solutions to key issues
| Editor's Note: This clause contains a table mapping between key issues and solutions.
Table 6.0-1: Mapping of solutions to key issues
Solutions
KI#1
KI#2
KI#3
KI#4
#1
X
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 6.1 Solution #1: reusing NDS/IP to N2 and Xn interfaces in WAB
| |
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 6.1.1 Introduction
| This solution proposes a the credential is provided to the WAB by OAM in the phase 2-1 of the WAB-node integration procedure defined in TS 38.401 [3]
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 6.1.2 Solution details
|
Figure 6.1.2-1 Procedure to configure the credential for NDS/IP connection
0. The WAB-node is pre-configured a credential for accessing to the OAM of WAB
Phase 1. WAB-MT Setup. It is described in TS 38.401[3].
Phase 2-1. WAB-gNB initialization. Addition to the description in TS 38.401[3], the WAB-gNB uses the pre-configured credential accessing to the OAM of WAB for authentication and security establishment. Then, the OAM of WAB sends the configuration data to the WAB-gNB in the secure link. The configuration data includes the credentials used for establishing Xn and N2 connections for UE. If the WAB servers UEs from different PLMN, the credentials may further bind with PLMN ID information.
Phase 2-2. WAB-gNB NG connection setup. Addition to the description in TS 38.401[3], the WAB-gNB uses the credential sent in Phase 2-1 to establish NDS/IP with the potential serving UE’s 5GC. If the WAB-gNB servers more than one PLMN, the WAB-gNB will use the corresponding credentials to establish NDS/IP with each UE’s PLMN’s 5GC.
Phase 2-3. WAB-gNB Xn connection setup. Addition to the description in TS 38.401[3], the WAB-gNB uses uses the credential sent in Phase 2-1 to establish NDS/IP with the potential serving UE’s NG-RAN. If the WAB-gNB servers more than one PLMN, the WAB-gNB will use the corresponding credentials to establish NDS/IP with each UE’s PLMN’s NG RAN.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 6.1.3 Evaluation
| The solution addresses the situation when the credential of UE’s 5GC or NG-RAN cannot be pre-configured at WAB. The phase 2-1 can be used to configure the credentials of potential serving UEs’ PLMN.
6.Y Solution #Y: <Solution Name>
6.Y.1 Introduction
Editor’s Note: Each solution should list the key issues being addressed.
6.Y.2 Solution details
6.Y.3 Evaluation
Editor’s Note: Each solution should motivate how the potential security requirements of the key issues being addressed are fulfilled.
|
98de5ba8eb100652bf114945d3c0bf11 | 33.724 | 7 Conclusions
| Editor’s Note: This clause contains the agreed conclusions that will form the basis for any normative work.
Annex <C>:
Change History
Change history
Date
Meeting
TDoc
CR
Rev
Cat
Subject/Comment
New version
2025-08
SA3#123
S3-252987
S3-252684 and S3-252686 for endorsed TR Skeleton for WAB Security
0.0.0
2025-10
SA3#124
S3-253741
Included changed from S3-253411, S3-253412, S3-253464, S3-253626, S3-253820 and S3-253821
0.1.0
2025-11
SA3#125
S3-254544
KI updated from S3-254168
0.2.0
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 1 Scope
| The present document …
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 2 References
| The following documents contain provisions which, through reference in this text, constitute provisions of the present document.
- References are either specific (identified by date of publication, edition number, version number, etc.) or non‑specific.
- For a specific reference, subsequent revisions do not apply.
- For a non-specific reference, the latest version applies. In the case of a reference to a 3GPP document (including a GSM document), a non-specific reference implicitly refers to the latest version of that document in the same Release as the present document.
[1] 3GPP TR 21.905: "Vocabulary for 3GPP Specifications".
[2] 3GPP TS 23.401: "General Packet Radio Service (GPRS) enhancements for Evolved Universal Terrestrial Radio Access Network (E-UTRAN) access".
[3] 3GPP TS 33.401: "3GPP System Architecture Evolution: Security Architecture".
[4] 3GPP TS 24.301: " Non-Access-Stratum (NAS) protocol for Evolved Packet System (EPS); Stage 3".
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 3 Definitions of terms and abbreviations
| |
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 3.1 Terms
| For the purposes of the present document, the terms given in 3GPP TR 21.905 [1] and the following apply. A term defined in the present document takes precedence over the definition of the same term, if any, in 3GPP TR 21.905 [1].
example: text used to clarify abstract rules by applying them literally.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 3.2 Abbreviations
| For the purposes of the present document, the abbreviations given in 3GPP TR 21.905 [1] and the following apply. An abbreviation defined in the present document takes precedence over the definition of the same abbreviation, if any, in 3GPP TR 21.905 [1].
<ABBREVIATION> <Expansion>
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 4 Architecture assumptions
| The following architecture assumptions are applied to the study:
- The general features and the Split MME architecture of Store and Forward Satellite operation are described in Annex O.2 of TS 23.401 [2] are used as architecture assumptions in this study.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 5 Key issues
| Editor’s Note: This clause contains all the key issues identified during the study.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 5.1 Key Issue #1: Authenticated UE to exchange NAS messages with multiple satellites in split-MME architecture
| |
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 5.1.1 Key issue details
| One of the architectural assumptions for Store and Forward Satellite operation is that when the service link is available, there is no feeder link and inter satellite link. There are two example deployment options for Store and Forward Satellite operation given in Annex O of TS 23.401 [2], i.e. Split MME architecture and Full EPC in each satellite.
For the split-MME architecture, S&F Satellite operation may involve multiple satellites allocated by an S&F Monitoring List. In this scenario, the UE context needs to be synchronized between the multiple MME-onboard(s) and the associated MME-ground. The synchronization of UE context between the MME-ground and MME-onboard(s) is out of the scope of 3GPP.
According to Annex N of TS 33.401 [3], regular LTE procedures are used to provide security between UE and network for the split-MME architecture. This means that once the UE completes an interaction with a satellite, the UE context in the satellite must be synchronized to other satellites before these satellites can perform any subsequent S&F Satellite operations with the UE. This significantly reduces the data exchange efficiency of the entire system.
Ideally, for an IoT device, once it is registered in the network and its UE context has been distributed to the satellites included in the S&F Monitoring List, the UE can exchange data with these satellites without the need for UE context synchronization between the satellites.
This key issue focuses on solutions that meet the following conditions:
- The UE context of the UE registered in the network has been provided to the satellites included in the S&F Monitoring List;
- The UE can perform Mobile Originated (MO) or Mobile Terminated (MT) data transmission with the satellites that have the UE context;
- The UE context does not need to be synchronized across the multiple satellites for supporting the MO/MT data transmissions. However, UE context synchronization may still be required for other changes not being associated with the MO/MT data transmission.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 5.1.2 Security threats
| If the NAS COUNTs are not synchronized across multiple satellites, an attacker may intercept and replay previously transmitted NAS messages. Since different satellites may accept outdated NAS COUNT values, the replay protection mechanism could be bypassed, leading to unauthorized actions.
Key stream may be reused if the security contexts are not well-managed across multiple satellites.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 5.1.3 Potential security requirements
| The 3GPP system shall support means to secure NAS messages exchange in the store and forward satellite operations.
5.X Key Issue #X: <Key Issue Name>
5.X.1 Key issue details
5.X.2 Security threats
5.X.3 Potential security requirements
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6 Solutions
| Editor’s Note: This clause contains the proposed solutions addressing the identified key issues.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.0 Mapping of Solutions to Key Issues
| Table 6.0-1: Mapping of Solutions to Key Issues
Key Issues
Solutions
1
1
X
2
X
3
X
4
X
5
X
6
X
7
X
8
X
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.1 Solution #1: Derivation of Satellite-Specific NAS keys for S&F Operation
| |
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.1.1 Introduction
| This solution addresses Key Issue #1: Authenticated UE to exchange NAS messages with multiple satellites in split-MME architecture.
This solution proposes a mechanism to derive unique NAS integrity and encryption keys for each satellite by using the satellite ID as an additional input parameter during the NAS key derivation.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.1.2 Solution details
| In this solution, it is proposed to derive distinct set of NAS keys for each satellite from the common root key KASME. The satellite-specific NAS keys are derived by the UE and the network using the KDF as specified in TS 33.220 [x].
For a serving Satellite n, the NAS integrity key KNASint and the NAS encryption key KNASenc are derived from the KASME with the following parameters as input:
- FC = 0xxx
- P0 = algorithm type distinguisher
- L0 = length of algorithm type distinguisher (i.e. 0x00 0x01)
- P1 = algorithm identity
- L1 = length of algorithm identity (i.e. 0x00 0x01)
- P2 = Satellite ID n.
- L2: length of Satellite ID n.
Where Satellite ID is an identifier uniquely indicating an MME-onboard. The Satellite ID of a given satellite is broadcast by the eNB within the SIB31 and the Satellite ID of the satellites that might be serving a given UE are included within the S&F Monitoring List, which is sent by the MME to indicate the satellite(s) that the UE may (re)-attempt NAS procedures (TS 23.401 clause 4.13.9.1).
As a result of using satellite-specific keys, the UE and each MME-onboard maintain independent pairs of NAS COUNT for their mutual communication. The NAS COUNTs are not synchronized with other satellites.
Upon receiving the UE context, the MME-onboard derives its satellite-specific NAS keys using its own Satellite ID. The MME-onboard can provide the UE an indicator, indicating that the separate NAS keys are implemented, together with the monitoring list. The UE derives the NAS keys for a Satellite ID immediately upon receiving the S&F Monitoring List, or when it determines to initiate a NAS procedure with a satellite present in the list with which no prior keys have been derived.
The UE and the MME-ground need to manage the multiple satellite-specific security context by associating each Satellite ID with the satellite-specific NAS keys and NAS COUNTs. When the MME-ground detects that the NAS COUNT for any satellite is about to wrap around, the MME-ground performs AKA procedure with the UE via any MME-onboard. Upon successful AKA completion, a new KASME is established. The MME-grounds then provide the new KASME to all MME-onboards in the S&F Monitoring List. The MME-onboards and the UE then derive new satellite-specific NAS keys and reset NAS COUNTs.
Editor’s Note: The establishment of AS security in this solution is FFS.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.1.3 Evaluation
| This solution proposed to use satellite-specific NAS keys for each satellite to prevent key stream reuse. There is no need to synchronize the NAS COUNT between satellites.
The solution has the following impacts:
- a new KDF needs to be defined;
- The UE and the MME-onboard needs to derive and store satellite-specific NAS keys;
- The UE and MME-ground needs to manage the multiple satellite-specific NAS keys and NAS COUNTs.
- The MME-onboard needs to provide the UE an indicator, indicating that the separate NAS keys are implemented, together with the monitoring list.
Editor’s Note: evaluations for the indicator is FFS.
- The MME-ground needs to provide updated KASME to all MME-onboards in the S&F Monitoring List.
- If multiple satellites are involved, multiple NAS security contexts need to be established between UE and MME.
- This solution is not backward compatible for pre-Rel-19 UEs.
Editor's Note: Further evaluation is FFS.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.2 Solution #2: NAS Security Context Isolation via Satellite-Specific NAS COUNT
| |
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.2.1 Introduction
| This solution addresses Key Issue #1: Authenticated UE to exchange NAS messages with multiple satellites in split-MME architecture.
This solution proposes a mechanism ensuing different satellite using different COUNT to protect NAS message and therefore eliminates the need for real-time NAS COUNT synchronization across satellites.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.2.2 Solution details
| This solution is based on the following assumptions and principles:
- the UE and each MME-onboard maintain independent pairs of NAS COUNTs (one for uplink, one for downlink) for their mutual communication. The NAS COUNTs are not synchronized with other satellites.
Based on the above principle, the existing procedures are reused to protect the NAS message between the UE and the network. The NAS integrity and confidentiality protection algorithms are same as defined in TS 33.401 [3], with the following modification to the construction of the 32-bit COUNT input parameter:
For a serving Satellite n:
COUNT := Satellite ID n || NAS OVERFLOW || NAS SQN
Where
- Satellite ID n is the 8-bit ID of Satellite n which is an identifier uniquely indicating an MME-onboard coded as a binary coded integer value from 0 to 255 as specified in 3GPP TS 24.301 [4]. The SatelliteID identifier of a given satellite is broadcast by the eNB within the SIB31 and the SatelliteID identifiers of the satellites that might be serving a given UE are included within the S&F Monitoring List, which is sent by the MME to indicate the satellite(s) that the UE may (re)-attempt NAS procedures (TS 23.401 clause 4.13.9.1)
- NAS OVERFLOW is a 16-bit value which is incremented each time the NAS SQN is incremented from the maximum value. It is maintained for the connection with Satellite n.
- NAS SQN is the 8-bit sequence number carried within each NAS message between UE and MME-onboard n. It is maintained for the connection with Satellite n.
All other input parameters (KEY=KNASint/KNASenc, BEARER, DIRECTION, LENGTH) and the algorithm execution remain unchanged.
The MME-onboard can provide the UE an indicator, indicating that the separate NAS COUNTs are implemented, together with the monitoring list. After receiving the indicator, the UE knows that the NAS COUNT needs to be construct with the above method.
The UE and the MME-ground need to manage the multiple satellite-specific security context by associating each Satellite ID with the satellite-specific NAS COUNTs. When the MME-ground detects that the NAS COUNT for any satellite is about to wrap around, the MME-ground performs AKA procedure with the UE via any MME-onboard. Upon successful AKA completion, a new KASME is established. The MME-grounds then provide the new KASME to all MME-onboards in the S&F Monitoring List. The MME-onboards and the UE then derive new NAS keys by running NAS SMC, and reset NAS COUNTs.
Editor's Note: The establishment of AS security in this solution is FFS.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.2.3 Evaluation
| This solution proposed to use satellite-specific NAS COUNTs for each satellite to prevent key stream reuse. There is no need to synchronize the NAS COUNT between satellites, and no change to the NAS keys.
The solution has the following impacts:
- A new NAS COUNT construction mechanism is needed;
- The UE and MME-ground needs to manage the multiple satellite-specific NAS COUNTs.
- The MME-onboard needs to provide the UE an indicator, indicating that the separate NAS keys are implemented, together with the monitoring list.
Editor’s Note: evaluations for the indicator is FFS.
- The MME-ground needs to provide updated KASME to all MME-onboards in the S&F Monitoring List.
- If multiple satellites are involved, multiple NAS security contexts need to be established between UE and MME.
- This solution is not backward compatible for pre-Rel-19 UEs.
Editor’s Note: Further evaluation is FFS.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.3 Solution #3: UE context management for S&F operation
| |
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.3.1 Introduction
| This solution addresses Key Issue #1: Authenticated UE to exchange NAS messages with multiple satellites in split-MME architecture.
After the UE is authenticated and NAS security is established, the satellite will send a security token to the UE, which contains the UE's current context. When the UE attempts to connect to another satellite, it will provide the security token to that satellite. The satellite will use the content in the security token to reconstruct the UE context and communicate directly with the UE through secure NAS messages.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.3.2 Solution details
| UE context management procedure for S&F operation is shown in the following figure.
Figure 6.3.2-1: UE context management procedure for S&F operation
0. The security key materials used to provide confidentiality and integrity protection for security tokens used in S&F operations are pre-configured in the satellites. The security tokens are used to transfer UE contexts from one satellite to another satellite.
1. The MME-ground provides UE authentication vectors to the MME-onboards when the feeder link is available.
2. The UE and satellite perform the authentication procedure when the service link is available.
3. The UE and satellite execute the Security Mode Command (SMC) procedure after the authentication procedure.
4. The UE and satellite exchange downlink/uplink data through secure NAS messages.
5. The satellite generates a security token based on the current context of the UE, which is protected by confidentiality and integrity using the security materials received in step 0. The token contains all the information required to reconstruct the UE context in another satellite, such as NAS keys, NAS COUNTs, etc.
NOTE 1: The detailed information of the security token structure and protection mechanism will be specified during the normative phase.
6. The satellite sends the security token to the UE through a NAS message and ends the connection with the UE. The satellite does not need to store the UE context after ending the connection with the UE. The UE stores the received security token.
7. When the UE connects to another satellite, it sends an Initial NAS message to the satellite, which includes the security token.
8. The satellite decrypts and verifies the security token using the security materials received in step 0. If the verification is successful, the satellite will attempt to exchange downlink/uplink data directly through secure NAS messages.
NOTE 2: Attackers may eavesdrop on communication between UEs and satellites, record the tokens, and then resend them to other satellites. Because the tokens are encrypted, attackers are unable to successfully perform subsequent NAS communication with the satellites. However, DoS attacks caused by this cannot be avoided.
9. The satellite performs the same operation as step 5
10. The satellite and UE performs the same operations as step 6.
11. The MME-onboards and MME-ground exchange downlink/uplink data when the feeder link is available.
Editor’s Note: The establishment of AS security in this solution is FFS.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.3.3 Evaluation
| This solution uses the target UE as an intermediate entity to securely transmit the UE context from one satellite to another, thereby meeting the requirements of Key Issue #1.
The advantages of this method are:
- UE can connect to any satellite that supports S&F services.
The disadvantages of this method are:
- Need to specify security token.
- Existing authentication and NAS security procedures need to be enhanced to include security token sharing from network to UEs.
- Security tokens add complexity to the security procedures.
- This solution is not backward compatible for pre-Rel-19 UEs.
- The security token is susceptible to a replay attack and could enable a DoS attack against the satellite.
- Transmission of the security token will increase the latency and probability of failure for NAS messages that include the token.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.4 Solution #4: Separate NAS COUNT pair per SatelliteID within an EPS Security Context
| |
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.4.1 Introduction
| This solution addresses Key Issue #1.
This solution is based on using separate pairs of NAS counters per Satellite ID in the EPS security context when the UE is served by multiple satellites operating in S&F mode and the UE registration remains valid even the serving satellite changes over time (i.e., the UE is not required to attach/detach in each satellite pass). The list of SatelliteID(s) in which the registration is valid is provided to the UE using the S&F Monitoring List.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.4.2 Solution details
| This solution applies to a satellite network operating in S&F mode and, it’s especially relevant for deployments based on the split MME architecture (see TS 23.402 Annex O.2) in which a UE registration remains valid across multiple satellites (unlike a full EPC deployment, where registration is only valid in one satellite).
The solution consists of enabling an option for the UE to use separate pairs of NAS counters (i.e. UL_NAS_Count and DL_NAS_Count) per SatelliteID within its EPS security context, where:
• SatelliteID is an identifier uniquely indicating an MME-onboard. The SatelliteID identifier of a given satellite is broadcast by the eNB within the SIB31 and the SatelliteID identifiers of the satellites that might be serving a given UE are included within the S&F Monitoring List, which is sent by the MME to indicate the satellite(s) that the UE may (re)-attempt NAS procedures (TS 23.401 clause 4.13.9.1)
• UL_NAS_Count is the uplink NAS counter related to the uplink NAS messages sent to the MME-onboard associated with SatelliteID.
• DL_NAS_Count is the downlink NAS counter related to the downlink NAS messages received from the MME-onboard associated with SatelliteID.
On the network side, this solution allows each MME-onboard to independently maintain its own pair of NAS counters, which shall no longer to be synchronised across the subset of the MME-onboard instances (identified each by a SatelliteID) that belong to the same logical MME in charge of the registered UE. This is depicted in Figure 6.Y.2-1, which is based on Figure O.2-1: "Split-MME" architecture for supporting Store and Forward Satellite operation for SMS and CP CIoT services” in Annex O.2 in TS 23.401.
Figure 6.4.2-1: Illustration of the solution consisting on using separate NAS COUNT pairs per SatelliteID
To ensure backward compatibility with Rel-19 UEs, which will still assume that NAS counters are synchronised across the satellites of the S&F Monitoring List, the proposed solution can be introduced as an optional capability for both UE and network (NW). Therefore, a UE capable of handling separate NAS counters per SatelliteID is expected to indicate such capability to the NW and the network , if capable of handling separate NAS counters per SatelliteID, should be able to indicate the UE whether this option is activated (i.e. the UE should use separate NAS counters per SatelliteID) or deactivated (i.e. the UE shall assume NAS counters are kept synchronised). In case the NW does not support this capability, the UE shall assume that NAS counters are synchronised across the satellites of the S&F Monitoring List.
This solution also considers the use of the “SatelliteID” value as part of the NAS COUNT 32-bit value.In this respect, the padding bits of the NAS Count are filled with the SatelliteID, as illustrated in Figure 2, so thatthe NAS messages used between the UE and a given satellite cannot be replayed with another satellite given NAS COUNT values will not match.
Figure 6.4.2-2: Filling NAS COUNT padding bits with SatelliteID
The activation of the EPS security context between the UE and the set of satellites of the S&F Monitoring List relies on:
(1) legacy LTE procedures for EPS security context activation as stated Annex N of TS 33.401, which is conducted between the UE and one of the MME-onboard entities, and,
(2) propagation/synchronisation of the activated EPS security context to the set of MME-onboard(s) and associated MME-ground, considering that how MME-onboard(s) interacts with MME-ground and how synchronization of the UE context between them is done is outside the scope of 3GPP, as stated in the principles of the split-MME architecture in 23.401 Annex O.2.
Accordingly, once the EPS security context is ready /synchronised in an MME-onboard, the UE and the MME-onboard can start interacting using a separate pair of NAS COUNTs. NAS COUNTs are assumed to be pre-set to zero in each MME-onboard(s).
Before any of the pairs of NAS COUNT associated with each of the satellites wraps around, the MME is expected to trigger EPS AKA run to activate fresh NAS keys. This will result in an update of the EPS security context (with the fresh NAS key and reset of NAS COUNT values) that shall be propagated/synchronised across all MME-onboard(s) and associated MME-ground in the same way that this is done when the EPS security context is established for the first time.
Editor’s Note: The establishment of AS security, handover procedure and cell reselection procedure in this solution is FFS.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.4.3 Evaluation
| The following impacts are needed:
- The EPS Security Context in the UE and MME needs to handle separate pairs of NAS COUNT per SatelliteID.
- A new NAS COUNT construction mechanism is needed to include the SatelliteID.
- To ensure backward compatibility, a new network capability and UE capability are needed to indicate support of satellite-specific NAS COUNTs.
The table below indicates how the security threats identified for KI#1 are accounted by the proposed solution.
Security threats (Section 5.2.2)
Mitigation
If the NAS COUNTs are not synchronized across multiple satellites, an attacker may intercept and replay previously transmitted NAS messages. Since different satellites may accept outdated NAS COUNT values, the replay protection mechanism could be bypassed, leading to unauthorized actions.
The NAS COUNT value includes the SatelliteID. This prevents reusing NAS COUNT values across the satellites.
Key stream may be reused if the security contexts are not well-managed across multiple satellites.
EPS security context across satellites are assumed to be always synchronized, excepting the NAS COUNT values which are independent for each satellite.
NAS COUNT values include the SatelliteID. This prevents reusing NAS COUNT values across the satellites and so reusing a key stream.
Editor's Note: Further evaluation is FFS.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.5 Solution #5: Protection for NAS message of authenticated UE in split-MME architecture
| |
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.5.1 Introduction
| This solution is proposed to address Key Issue #1, providing a protection method for exchanging the NAS message in the Store and Forward satellite operations.
As specified in TS 33.401 [3], the NAS security is terminated on the MME-onboard, and the ground segment of the network ensures that the latest NAS security context of the UE is available at the MME-onboard. When multiple satellites are involved in the Store and Forward satellite operation, the NAS COUNTs should be synchronized to mitigate the replay attack.
This solution proposes that NAS COUNTs are maintained and managed by the UE and MME-ground. When a DL NAS message of authenticated UE is received, the MME-ground is responsible for selecting the MME on-board based on the coverage availability information. As defined in TS 23.401 [2], the satellite coverage availability information provisioned to the MME describes when and where satellite coverage with both service link and feeder link connectivity is expected or not expected to be available in an area. By using the coverage availability information, this solution assumes that the UE can receive the DL NAS messages from MME on-board(s) in sequence.
In other words, the MME-ground selects the MME on-board that will be available to the UE earliest. For the selected MME on-board, the MME-ground provides the value of DL NAS COUNT together with the DL NAS signaling. Since the selection is based on the coverage availability information, the MME on-board(s) will be available for UE in sequence and the value of DL NAS COUNT will be received in order, which mitigates the replay attack in the Store and Forward satellite operations.
For UL NAS messages of authenticated UE, the UE includes the Satellite ID in the UL NAS signalling, then uses the NAS security keys to protect the UL NAS signalling, including the Satellite ID. Once receiving the NAS signalling, the MME on-board verifies the integrity by using the NAS security key. If the verification is successful, the MME on-board further checks whether the received Satellite ID is associated with the onboard satellite. By checking the Satellite ID, the replay attack (i.e., resend the protected NAS signaling to another MME on-board) can be detected.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.5.2 Solution details
| 6.5.2.1 DL NAS signalling protection
Figure 6.5.2-1: Protection for DL NAS messages of authenticated UE
0. The UE and MME-ground hold the latest NAS COUNTs, including the UL NAS COUNT and DL NAS COUNT.
At Time 1:
1. The MME-ground receives the DL NAS signaling #1 of the authenticated UE from another EPS NF.
2. Based on the coverage availability information, the MME-ground selects one of the MME on-board(s) (e.g. MME on-board the SAT1) to transmit the DL NAS signaling #1.
3. The MME-ground sends the DL NAS signaling #1 together with the latest value of DL NAS COUNT (e.g. DL NAS COUNT #1), and increases the DL NAS COUNT by one.
If the service link is not available, the MME on-board the SAT1 stores the DL NAS COUNT #1 together with the DL NAS signaling #1.
At Time 2:
4. The MME-ground receives the DL NAS signaling #2 of the authenticated UE from another EPS NF.
5. Based on the coverage availability information, the MME-ground selects one of the MME on-board(s) (e.g. MME on-board the SAT2) to transmit the DL NAS signaling #1.
6. The MME-ground sends the DL NAS signaling #2 together with the latest value of DL NAS COUNT (e.g. DL NAS COUNT #2), and increases the DL NAS COUNT by one.
If the service link is not available, the MME on-board the SAT2 stores the DL NAS COUNT #2 together with the DL NAS signaling #2.
At Time 3 and Time 4, the UE can receive the protected DL NAS message in sequence.
7. Once the service link becomes available (Time 3), the MME on-board the SAT1 generates the integrity-protected and confidentiality-protected NAS signaling #1 and sends it to the UE.
NOTE 1: Time 3 may happen before Time 2. In this case, Step #7 is performed before Steps #4-6.
8. Once the service link becomes available (Time 4), the MME on-board the SAT2 generates the integrity-protected and confidentiality-protected NAS signaling #2 and sends it to the UE.
Editor’s Note: How to deal with a scenario where DL NAS messages are delivered out-of-order is FFS.
Editor’s Note: The establishment of AS security in this solution is FFS.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.5.2.2 UL NAS signalling protection
| In the split-MME architecture, the UE includes the Satellite ID in the UL NAS signalling, then uses the NAS security keys to protect the UL NAS signalling, including the Satellite ID.
Once receiving the NAS signalling, the MME on-board verifies the integrity by using the NAS security key. If the verification is successful, the MME on-board further checks whether the received Satellite ID is associated with the onboard satellite. If the received Satellite ID matches with the identifier of satellite hosting the MME on-board, the MME on-board stores the UL NAS DATA and updates the local stored UE NAS context (i.e., increases the UL NAS COUNT). Otherwise, the MME on-board discards this NAS signalling.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.5.3 Evaluation
| This solution addresses the security requirements of Key Issue #1.
For the protection of DL NAS messages, the coverage availability information is used by the MME-ground for selecting the MME on-board. By using the coverage availability information, this solution assumes that the UE can receive the DL NAS messages from MME on-board(s) in sequence.
For the protection of UL NAS messages, the Satellite ID is included in the NAS signalling and protected by the NAS security keys.
This solution is aligned with the security mechanism defined in Annex N of TS 33.401 [3] and UE security handling for EPS systems.
Editor’s Note: Further evaluation is FFS.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.6 Solution #6: Secure NAS messages via using different NAS keys in multiple satellites
| |
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.6.1 Introduction
| This solution addresses “Key issue #1: Authenticated UE to exchange NAS messages with multiple satellites in split-MME architecture”.
This solution is based on split MME architecture. S&F Satellite operation may involve multiple satellites allocated by an S&F Monitoring List. In order to prevent reusing key stream, one possible approach is to use different NAS keys when UE interacts with different satellites. This solution can improve the data exchange efficiency of the entire system.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.6.2 Solution details
| Based on the existing authentication procedures, this solution proposes to use different NAS keys when UE exchanges data with multiple satellites.
Figure 6.6.2-1 Enhanced NAS security for multiple satellites in S&F mode
SAT#1 has available Service Link.
1. The UE sends the Attach Request to SAT#1.
2. If SAT#1 does not have context to authenticate the UE, then sends the Attach Reject.
SAT#1 has available Feeder Link.
3. SAT#1 sends the Attach Request to the MME-ground.
4. The MME-ground obtains authentication data including KASME, as defined in TS 33.401 [3].
5. The MME-ground determines to use SAT#1 to serve UE, then the MME-ground calculates KASME1* by using KASME and SAT Id of SAT#1.
6. The MME-ground distributes KASME1* for SAT#1 during the transmission of AV.
SAT#1 has available Service Link.
7. The authentication procedure is completed, as defined in TS 33.401 [3].
8. SAT#1 derives NAS keys based on the KASME1* using existing mechanism as defined in TS 33.401[3] and sends the NAS security mode command integrity protected.
9. The UE calculates KASME1* using the same method as the MME-ground in step5, and further derives the NAS keys using existing mechanism as defined in TS 33.401[3], then the UE verifies the NAS security mode command.
10. If successfully verified, the UE sends the NAS security mode complete to SAT#1.
11. After the NAS SMC procedure, the UE and SAT#1 send protected NAS messages.
SAT#2 has available Feeder Link.
12. The MME-ground determines to use SAT#2 to serve the UE, the MME-ground calculates KASME2* by using KASME and SAT Id of SAT#2.
13. The MME-ground distributes KASME2* for SAT#2. Then SAT#2 derives the NAS keys by using KASME2*.
SAT#2 has available Service Link.
14. The UE establishes RRC connection with SAT#2 and calculates KASME2* using the method as the MME-ground in step12, and further derives the NAS keys by using KASME2*.
NOTE 1: The UE obtains the SAT Id of SAT#2 broadcast by the satellite SAT#2 to derive KASME2*.
The UE sends a protected initial NAS message to SAT#2. The successful processing of the initial NAS message by the satellite SAT#2 activates the NAS key between the UE and the satellite SAT#2. The UE and SAT#2 exchange protected NAS messages.
NOTE 2: As described in TS 23.401[2], the MME-ground together with the associated MME-onboard(s) behave jointly as a single MME entity. For multiple satellites, assume MME-onboards have the same list of ordered NAS security algorithms. After NAS SMC, the selected NAS security algorithms could be synchronized for MME-onboards. Each satellite/UE pair maintains independent COUNTs.
NOTE 3: As Store-and-Forward satellite operations involve sparse, infrequent transmissions, spread across multiple satellites, COUNT wrap-around can only occur over a substantial amount of time. It is left to network policy to determine when the MME ground re-authenticates the UE to avoid COUNT wrap around.
NOTE 4: To secure NAS messages using different NAS keys during handover-like processes in Store-and-Forward satellite operations, the UE handles satellite changes as cell reselection events and performs a protected tracking area update procedure using NAS keys.
Editor’s note: The establishment of AS security in this solution is FFS.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.6.3 Evaluation
| This solution addresses the Key Issue #1, and it applies for S&F operations with multiple satellites. In this solution, the UE can exchange data with multiple satellites efficiently without security risk.
The solution has the following impacts:
This solution requires the MME-ground and the UE to derive new keys (i.e. KASME* derivation based on KASME) for different satellites and the UE needs to maintain multiple NAS COUNTs. This solution requires the key transfer from MME-ground to MME-onboard.
If the UE interacts with a new satellite, it computes a new KASME* and derives new NAS keys based on the same NAS security algorithms.
The same UE context cannot be re-used across satellites because new NAS keys need to be generated by the UE.
Each MME on-board needs to maintain a separate pair of NAS COUNTs.
Editor’s Note: The detail on securing NAS messages using different NAS keys during handover-like process is FFS.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.7 Solution #7: Solution for NAS COUNT synchronization in store-and-forward operations
| |
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.7.1 Introduction
| As per the threat described in the key issue #1, an attacker may intercept and replay previously transmitted NAS messages. This solution proposes the following to address this threat:
• A new “Satellite access information” can be included as part of Initial UE message sent from satellite eNB to MME. This information can be used by MME to enable UE context synchronization including NAS COUNT verification and synchronization for the satellites included in the S&F Monitoring List.
◦ A “3GPP satellite access type” in Access type information element (reference : TS 24.501 [X] clause 9.11.2.1A) is included. Considering satellite access as a different access type to enable an independent NAS COUNT for “3GPP satellite access type”.
• MME-onboard and MME-onground synchronize the NAS COUNT values for UEs whose security contexts are provided to the satellites included in the S&F Monitoring List. The mechanism of this synchronization across multiple satellites is out of 3GPP scope, however, 3GPP can recommend certain actions as follows:
◦ MME-onboard and MME-onground need to ensure that a given NAS COUNT value shall be accepted at most one time and only if message integrity verifies correctly. This is in accordance with clause 4.4.3.2 from TS 24.501 [X].
◦ If MME-onboard receives a new message from a UE for which the UE security context is available with the satellite, and the integrity verification is verified successfully, the MME-onboard:
▪ Request MME-onground for NAS COUNT duplicate verification. This can also be done using NAS sequence number verification.
▪ If MME-onground responds indicating that the NAS COUNT is duplicate, OR if there is a timeout because of long delay in obtaining the feeder link, MME-onboard discards that message from UE.
▪ If MME-onground responds indicating that the NAS COUNT is NOT duplicate, MME-onboard consider it as a valid message and proceed to ensure seamless connectivity for the UE.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.7.2 Solution details
|
Figure 6.7.2-1: Message sequence showing NAS COUNT verification at MME
As shown in Figure 6.7.2-1:
- In Step 2, UL and DL NAS COUNTs are synchronized between MME-onboard and MME-onground entities. Note that from UE’s perspective, MME is expected to be seen as a single logical entity. Hence, in this solution, the proposal is to ensure UL and DL NAS COUNT synchronization between MME-onboard entities to ensure replay protection.
- In Step 3, if a genuine UE sends a NAS message, with UE security context available in Satellite#2, the integrity verification succeeds. The MME-onboard stores the message in the UE security context.
- In Step 5, MME-onboard requests the NAS COUNT verification with MME-onground.
- In Step 6, MME-onground responds with the verification status.
When messages are received simultaneously from multiple satellites by MME-onground, then the coordination between MME-onboards and MME-onground ensures that duplicates are dropped and the NAS security context is maintained seamlessly for the UE.
- In case feeder link is not available for a long time, and there may be a timeout implemented, the MME-onboard drops this NAS message from the UE. Also, if the UL NAS COUNT verification status indicates duplicate or old NAS message, the MME-onboard drops it in order to ensure replay protection requirements stated in clause 4.4.3.2 of TS 24.501.
- If the UL NAS COUNT verification status from MME-onground indicates that it is not a duplicate or old message, MME-onboard processes it further.
- In Step 8, MME on-ground can provide latest DL NAS COUNT values to MME onboard of satellite 2 which is now the serving satellite for the UE. This step can be executed conditionally if the UL NAS COUNT verification succeeds.
Editor's Note: The establishment of AS Security in this solution is FFS.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.7.3 Evaluation
| TBD
Editor’s Note: The impact on signaling to mme on-ground needs to be noted.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.8 Solution #8: New specific rules to handle NAS Counter Overflow in S&F mode
| |
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.8.1 Introduction
| This solution addresses KI#1.
In S&F Satellite operation, the subset of satellites operating in S&F Mode in which a given UE registration is valid (i.e. satellites included in the S&F Monitoring List), are expected to maintain a synchronised UE context, even though the synchronisation mechanism is outside the scope of 3GPP.
This solution proposes to add an exception with respect to the synchronisation of the NAS counters. Based on the added exception, this solution proposes introducing specific rules for managing the pair of NAS counters stored by the UE and by the MME operating S&F Mode as follows:
• When a UE registration is valid in multiple satellites operating in S&F mode, each time the UE interacts with one of these satellites, the UL NAS Overflow Counter (OC) stored in the UE may be higher than the UL NAS OC stored in the MME of the serving satellite. This discrepancy can arise due to previous interactions between the UE and other serving satellites of the same PLMN, where the UL NAS OC in the UE was incremented due to the UL NAS SQN wrap-around. In such a case, if the MME fails to verify the integrity of a received NAS packet using the last stored UL NAS OC, the MME may attempt to validate the message integrity using a series of consecutively incremented UL NAS OC values. If one of the attempts is successful, the MME updates its stored UL NAS OC accordingly.
• Similarly, when a UE registration is valid in multiple satellites operating in S&F mode, the DL NAS Overflow Counter (OC) stored in the UE may be higher than the last DL NAS OC stored in the MME of the serving satellite. This can result from prior interactions between the UE and other satellites where the DL NAS OC in the UE was incremented due to the DL NAS SQN wrap-around. In such cases, if the UE fails to verify the integrity of a received NAS packet using the last stored DL NAS OC, it may attempt to validate the message integrity using a series of consecutively decremented DL NAS OC values. To avoid replay attack, the UE can rely on the fact that the Satellite ID is not the same.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.8.2 Solution details
| This section provides further details on this solution by analysing the uplink case (UE MME-onboard) and the downlink case (UE MME-onboard) when considering (1) UE is served by multiple satellites as per the S&F Monitoring List provided to the UE and (2) UE assumes that NAS counters in the MME-onboard(s) of those satellites are not necessarily synchronised.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.8.2.1 Uplink case
| Figure 6.8.2.1-1 shows the steps taken by the MME-onboard. Changes introduced by this solution are marked in red.
a) Upon receiving an integrity protected NAS uplink message, the MME-onboard retrieves the SQN and the NAS message authentication code (NAS-MAC) which are then used to compute the expected NAS message authentication code (XNAS-MAC) according to 3GPP TS 33.401 clause 8.1 and Annex B.2. Furthermore, the UL NAS Count is increased according to 3GPP TS 24.301 clause 4.4.3.
b) The computed XNAS-MAC is compared with the NAS-MAC received in the NAS integrity protected message.
c) If the two codes match up, the integrity check is successful and the MME-onboard can process the uplink NAS message.
d) If the two codes do not match up, the MME-onboard increases the UL OC by 1 which means increasing the UL NAS Count by 256 units.
e) The XNAS-MAC is computed again and compared with the NAS-MAC. If the two codes match up, step c) is executed. Otherwise, step d) is executed.
Note that steps d) and e) are repeated up to a number X of times. If the number of attempts exceeds X, the integrity check fails and the NAS message is discarded.
A successful integrity check also indicates that the UL NAS OC has been correctly estimated and that the UE and the MME-onboard are aligned, i.e., the UL NAS OC stored by the UE match the UL NAS OC stored by the MME-onboard.
Figure 6.8.2.1-1: Handling of UL NAS OC in the MME-onboard.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.8.2.2 Downlink case
| Figure 6.8.2.2-1 shows the steps taken by the UE. Changes introduced by this solution are marked in red.
a) Upon receiving an integrity protected NAS downlink message, the UE retrieves the SQN and the NAS message authentication code (NAS-MAC) which are then used to compute the expected NAS message authentication code (XNAS-MAC) according to 3GPP TS 33.401 clause 8.1 and Annex B.2. Furthermore, the DL NAS Count is increased according to 3GPP TS 24.301 clause 4.4.3.
b) The computed XNAS-MAC is compared with the NAS-MAC received in the NAS integrity protected message.
c) If the two codes match up, the integrity check is successful and the UE can process the downlink NAS message.
d) If the two codes do not match up, and the SatelliteID of the current serving satellite is different from the SatelliteID of the previous serving satellite, the UE decreases the UL OC by 1 which means decreasing the UL NAS Count by 256 units.
e) The XNAS-MAC is computed again and compared with the NAS-MAC. If the two codes match up, step c) is executed. Otherwise, step d) is executed.
Note that steps d) and e) are repeated up to a number X of times. If the number of attempts exceeds X, the integrity check fails and the NAS message is discarded.
A successful integrity check also indicates that the DL NAS OC has been correctly estimated and that the MME-onboard and the UE are aligned, i.e., the DL NAS OC stored by the MME-onboard matches the DL NAS OC stored by the UE.
SatelliteID is an identifier uniquely indicating an MME-onboard. The SatelliteID identifier of a given satellite is broadcast by the eNB within the SIB31 and the SatelliteID identifiers of the satellites that might be serving a given UE are included within the S&F Monitoring List, which is sent by the MME to indicate the satellite(s) that the UE may (re)-attempt NAS procedures (TS 23.401 clause 4.13.9.1).
Figure 6.8.2.2-1: Handling of DL NAS OC in the UE.
Editor’s Note: The impact of repeating integrity verification for X times on the UE side and MME-onboard (i.e., fake messages sent by the attacker require more resources for integrity verification, which increases the risk of attack) is FFS.
Editor’s Note: The impact of accepting a range of NAS counters to the overall security of 3GPP system is FFS
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 6.8.3 Evaluation
| Editor’s Note: Each solution should motivate how the potential security requirements of the key issues being addressed are fulfilled.
6.Y Solution #Y: <Solution Name>
6.Y.1 Introduction
Editor’s Note: Each solution should list the key issues being addressed.
6.Y.2 Solution details
6.Y.3 Evaluation
Editor’s Note: Each solution should motivate how the potential security requirements of the key issues being addressed are fulfilled.
|
6cba009aef02ddb1c451c3c3fe4d2645 | 33.700-30 | 7 Conclusions
| 7.Z Key Issue #Z: <Key Issue Name>
Editor’s Note: This clause contains the agreed conclusions of Key Issue #Z.
Annex <A>:
<Informative annex title for a Technical Report>
Annex <X>:
Change history
Change history
Date
Meeting
TDoc
CR
Rev
Cat
Subject/Comment
New version
2025-10
SA3#124
S3-253723
Incorporate TR skeleton, new Key Issue and new solutions
0.1.0
2025-11
SA3#125
S3-254540
Incorporate solution updates
0.2.0
|
c2dec694608a380e98d070db670d21f3 | 33.746 | 1 Scope
| The present document studies the potential security enhancements for 5G NR Femto. More specifically, the study investigates potential security enhancements in the following areas:
- The security requirements and potential solutions to enhance the security of NR Femto devices, to detect misconfigured or compromised NR Femto devices, and to eliminate the security impacts from misconfigured or compromised NR Femto devices.
- The security and privacy aspects of local access for NR Femto scenario.
|
c2dec694608a380e98d070db670d21f3 | 33.746 | 2 References
| The following documents contain provisions which, through reference in this text, constitute provisions of the present document.
- References are either specific (identified by date of publication, edition number, version number, etc.) or non‑specific.
- For a specific reference, subsequent revisions do not apply.
- For a non-specific reference, the latest version applies. In the case of a reference to a 3GPP document (including a GSM document), a non-specific reference implicitly refers to the latest version of that document in the same Release as the present document.
[1] 3GPP TR 21.905: "Vocabulary for 3GPP Specifications".
[2] 3GPP TS 23.501: "System architecture for the 5G System (5GS)".
[3] 3GPP TS 33.545: "Security aspects of NR Femto".
|
c2dec694608a380e98d070db670d21f3 | 33.746 | 3 Definitions of terms, symbols and abbreviations
| |
c2dec694608a380e98d070db670d21f3 | 33.746 | 3.1 Terms
| For the purposes of the present document, the terms given in TR 21.905 [1] and the following apply. A term defined in the present document takes precedence over the definition of the same term, if any, in TR 21.905 [1].
example: text used to clarify abstract rules by applying them literally.
|
c2dec694608a380e98d070db670d21f3 | 33.746 | 3.2 Symbols
| For the purposes of the present document, the following symbols apply:
<symbol> <Explanation>
|
c2dec694608a380e98d070db670d21f3 | 33.746 | 3.3 Abbreviations
| For the purposes of the present document, the abbreviations given in TR 21.905 [1] and the following apply. An abbreviation defined in the present document takes precedence over the definition of the same abbreviation, if any, in TR 21.905 [1].
<ABBREVIATION> <Expansion>
|
c2dec694608a380e98d070db670d21f3 | 33.746 | 4 Security Architecture and Assumptions
| The following security architecture and assumptions are applied to the present document:
- Annex V in TS 23.501[2] captures the architecture for NR Femto. The architecture option of NR Femto with a local UPF is reused as the basis for this study.
- The security architectural and requirements captured in TS 33.545 [3] is reused as basis for this study.
|
c2dec694608a380e98d070db670d21f3 | 33.746 | 5 Key issues
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.