hash
stringlengths 32
32
| doc_id
stringlengths 5
12
| section
stringlengths 4
1.47k
| content
stringlengths 0
6.67M
|
|---|---|---|---|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
1 Scope
|
This document studies the security architecture, i.e. the security features and the security mechanisms for inter-access mobility between 3GPP access system and non-3GPP access systems. For the general architecture for inter-access mobility cf. TR 23.882. This report is meant to provide more detail on the security aspects of inter-access mobility.
The scope should be extended to the mobility between two non-3GPP access systems, which interwork with 3GPP core entities. An example would be the mobility between two WLAN access systems providing 3GPP IP access.
Disclaimer: This TR reflects the discussions held in 3GPP SA3 while 3GPP SA3 was working towards TS 33.402 [14]. This TR may therefore be useful to better understand the basis on which decisions in TS 33.402 [14] were taken, and which alternatives were under discussion. However, none of the text in this TR shall be quoted as reflecting 3GPP’s position in any way. Rather, 3GPP’s position on security for non-3GPP access to EPS is reflected in the normative text in TS 33.402 [14]. Information in the TR may be inaccurate and outdated. One example of outdated text can be found in clauses 4.1 and 4.2 on alternatives for authentication protocols. The choices of authentication protocols finally made by 3GPP can be found in TS 33.401 [13] and TS 33.402 [14] respectively.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
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 23.882: "3rd Generation Partnership Project; 3GPP System Architecture Evolution: Report on Technical Options and Conclusions".
[2] 3GPP TS 33.234: "3rd Generation Partnership Project; Wireless Local Area Network (WLAN) interworking security".
[3] 3GPP TS 29.061: "3rd Generation Partnership Project; Technical Specification Group Core Network; Interworking between the Public Land Mobile Network (PLMN) supporting packet based services and Packet Data Networks (PDN)".
[4] 3GPP TS 33.210: "3G security; Network Domain Security (NDS); IP network layer security".
[5] “IKEv2 Mobility and Multihoming Protocol (MOBIKE)”, draft-ietf-mobike-protocol-03.txt, Sep 2005.
[6] RFC 3957 “Authentication, Authorization, and Accounting (AAA) Registration Keys for Mobile IPv4”.
[7] "NETLMM protocol", draft-giaretta-netlmm-dt-protocol-00.txt, June 2006.
[8] RFC 4285 “Authentication Protocol for Mobile IPv6”.
[9] “Mobile IPv6 Bootstrapping for the Authentication Option Protocol”, draft-devarapalli-mip6-authprotocol-bootstrap-03.txt, September 2007.
[10] “Diameter Mobile IPv6: Support for Home Agent to Diameter Server Interaction”, draft-ietf-dime-mip6-split-05.txt, September 2007.
[11] “Proxy Mobile IPv6”, draft-ietf-netlmm-proxymip6-06.txt, September 2007.
[12] RFC4832 “Security threats of network based mobility management”.
[13] 3GPP TS 33.401: "3GPP System Architecture Evolution (SAE); Security Architecture".
[14] 3GPP TS 33.402: "3GPP System Architecture Evolution (SAE); Security aspects of non- 3GPP accesses".
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
3 Definitions, symbols and abbreviations
| |
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
3.1 Definitions
|
For the purposes of the present document, the following apply:
Access network: one of following access network: GPRS IP access, WLAN 3GPP IP access, WLAN Direct IP access LTE, WiMax, etc.
Data origin authentication: The corroboration that the source of data received is as claimed.
WLAN 3GPP IP Access: Access to an IP network via the 3GPP system.
WLAN Direct IP Access: Access to an IP network is direct from the WLAN AN.
3GPP - WLAN Interworking: Used generically to refer to interworking between the 3GPP system and the WLAN family of standards.
Trusted Access: A non-3GPP IP Access Network is defined as a “trusted non-3GPP IP Access Network” if the 3GPP EPC system chooses to trust such non-3GPP IP access network. The 3GPP EPC system may choose to trust the non-3GPP IP access network operated by the same or different operators, e.g. based on business agreements. Specific security mechanisms may be in place between the trusted non-3GPP IP Access Network and the 3GPP EPC to avoid security threats. The decision whether a specific non-3GPP IP Access Network is trusted or untrusted is up to the 3GPP EPC operator, and is not based on the specific link-layer technology adopted by the non-3GPP IP Access Network.
Source access system: in handover situations, this is the access system, from which the UE is handed over.
Target access system: in handover situations, this is the access system, to which the UE is handed over.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
3.2 Symbols
|
For the purposes of the present document, the following symbols apply:
Gi Reference point between GPRS and an external packet data network
Wi Reference point is similar to the Gi reference point, applies to WLAN 3GPP IP Access
Wm Reference point is located between 3GPP AAA Server and Packet Data Gateway respectively between 3GPP AAA Proxy and Packet Data Gateway
Wu Reference point is located between the WLAN UE and the PDG. It represents the WLAN UE-initiated tunnel between the WLAN UE and the PDG
Gi+/Wi+ Mobile IP signalling and bearer plane between the Gateway (i.e. GGSN or PDG) and the MIP HA;
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
3.3 Abbreviations
|
For the purposes of the present document, the following abbreviations apply:
AAA Authentication Authorisation Accounting
AN Access network
APN Access Point Name
BSF Bootstrapping Function
DS-MIPv6 Dual stack MIP
FA Foreign Agent
GBA Generic Bootstrapping Architecture
GGSN Gateway GPRS Support Node
HA Home agency
HN Home network
IP Internet Protocol
IPSec IP Security protocol
I-WAN Interworking Wireless Local Area Network
MIP IP mobility
MOBIKE IKEv2 Mobility and Multihoming Protocol
MS Mobile Station
MN Mobile Node
NAI Network Access Identifier
NAT Network Address Translation
NAF Network Application Function
NETLMM Network-based localized mobility management
PDG Packet Data Gateway
PDP Packet Data Protocol
RFC Request For Comments
RRQ MIPv4 Registration Request
RRP MIPv4 Registration Response
SAE System Architecture Evolution
SGSN Serving GPRS Support Node
SPI Security Parameter Index
URI Uniform Resource Identifier
USIM UMTS subscriber identity module
UE User Equipment
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
4 Authentication protocols across access systems
|
Editor’s note: it will be decided later if this section is needed in the final report.
It is assumed that an SAE user has a USIM which is used as user credential in authentication.
Authentication protocols are assumed to be run between the UE and an authentication server in the home network. It is likely there will always be a 3G AAA server to terminate authentication protocols in SAE, but this is still to be decided by SA2 (i.e. it is still to be decided whether always AAA protocols, e.g. DIAMETER, will be used to carry authentication data, or whether MAP may still be used). When AKA is used then the 3G AAA server will interface with a 3G Authentication Centre.
Even for one user, the type of authentication protocol depends on the type of access network. E.g. for I-WLAN EAP-AKA may be used, whereas for UTRAN UMTS AKA will be used.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
4.1 UMTS AKA
|
UMTS AKA will be used across UTRAN. It is still to be decided by SA3 whether UMTS AKA or EAP-AKA will be used over LTE.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
4.2 EAP-AKA
|
EAP-AKA may be used across I-WLAN and for WiMAX.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
4.3 Others
| |
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
5 Establishment of security contexts in the target access system
|
Each type of access system may require there own security contexts, which may need to be available to protect the access network. An example is an MSK key in a WLAN access system using an EAP method for authentication and key agreement. The MSK is then used to derive further keys.
An example of an access system more complex than WLAN and requiring more security contexts to be set up is WiMAX. WiMAX does not only need keys for the protection of the link layer, but e.g. also keys to protect Mobile IP signalling of the WiMAX-internal Mobile IP (CMIP or PMIP) layer providing WiMAX-internal mobility, which is different from the SAE Mobile IP layer providing mobility between access systems, of which at least one is non-3GPP.
There may also be access systems, which do not require any security context, e.g. a DSL-based access system relying on physical security.
The establishment of these security contexts in the access system may be done in two ways:
with the support of SAE;
without the support of SAE.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
5.1 Establishment of security contexts with the support of SAE
|
In this case, the credentials the UE shares with the 3G AAA server are used to establish security contexts in the access system. An example of this case is I-WLAN Direct IP access, where the SIM or USIM are used to establish MSK required to protect the WLAN link layer. Another example is likely WiMAX: the WiMAX Forum is currently working on solutions for 3G-WiMAX interworking, which would allow to bootstrap WiMAX-internal security contexts from a key derived from a run of EAP-AKA between the UE and the 3G AAA server.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
5.2 Establishment of security contexts without the support of SAE
|
In this case, credentials other than those available in 3G networks are used to establish security contexts in the non-3GPP access system. An example of this case is WiMAX when WiMAX-specific credentials are used to set up IP connectivity across WiMAX. SAE plays no role in this set up, so the establishment of these security contexts is out of scope of SAE.
It is assumed that the SAE user always uses a USIM on UICC to perform mutual authentication and establish security contexts with the Home Network.
It is to be decided by SA3 whether a UE-PDG tunnel is required.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
6 Establishment of IPsec tunnel between UE and PDG across the target non-3GPP access system (if required)
|
One of the two variants of the S2 interface in the SAE architecture, cf. TR 23.882, allows to connect an access system to the evolved SAE packet core via an IPsec tunnel between the UE and a PDG. WLAN 3GPP IP access is an example of the use of such a tunnel, but WLAN is not the only access system which may be connected in this way. This section deals with the roaming of a UE between an access system (old) to another access system (new), for the case that at least the target access system requires such a UE-PDG tunnel.
The level of security achieved in certain deployments of non-3GPP IP access networks though internal security mechanisms (including confidentiality, integrity protection, protection of signalling, key management, etc) of some such non-3GPP IP access networks may be trusted by the 3GPP Evolved Packet Core (EPC) operator. In such case, no additional security mechanisms (e.g. IPSec tunnels from the UE to the EPC) are required. in the sense that the non-3GPP IP access network can interwork with the 3GPP EPC without relying on an IPsec tunnel to the UE. Such non-3GPP IP access networks are referred to here as "trusted non-3GPP IP access networks". The decision whether a specific non-3GPP IP access network is trusted or untrusted is up to the 3GPP EPC operator and is not based on the specific link-layer technology adopted by the non-3GPP IP access network.
If the non-3GPP IP access network is trusted (i.e. based on business, roaming and interconnection agreements), the need for a PDG functionality to connect the non-3GPP IP access to the EPC is FFS.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
6.1 The source access system has a UE-PDG tunnel
|
An example of this case is mobility between two I-WLAN 3GPP IP access systems. The problem to be solved is to retain the IPsec tunnel even when the IP address of the UE changes due to mobility.
There are two cases here: the PDG remains the same or the PDG changes.
If PDG remains the same, the existing IPsec tunnel could be maintained. In order to achieve this, a mechanism proposed in TR 23.882, Annex E, is MOBIKE. For MOBIKE to work, it is required that the PDG remains the same while the UE moves.
If the PDG changes, then it is not a matter of maintaining the IPSec tunnel, but creating a new one with the target PDG. In such case, the focus becomes the mechanisms on the S2 interface, not what happens between the new PDG and the UE.
Another possible solution to retain the IPsec tunnel when the PDG remains fixed would be the use of an IP mobility mechanism (e.g. Mobile IP). The Mobile IP Home Agent would have to be e.g. located between the PDG and the UE, but close to the PDG, ensuring that the outer IP address of the IPsec tunnel remains constant, even while the UE moves and acquires a new local IP address. The adoption of MIP for mobility is FFS.
If the PDG changes, then it is not a matter of maintaining the IPSec tunnel, but creating a new one with the target PDG. In such case, in addition, to the establishment of the new IPsec tunnel, the mobility of the PDG has to be handled by the S2 interface.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
6.2 The source access system does not have a UE-PDG tunnel
|
An example of this case is mobility between a 3GPP access system, such as LTE or UTRAN, and an I-WLAN 3GPP IP access system. The problem to be solved is to set up the IPsec tunnel in the target system in an efficient way.
Neither MOBIKE nor an additional layer of Mobile IP will help here.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7 Security for IP based mobility
|
There may be several layers of Mobile IP being used in a complete SAE system, including access networks. E.g. there is a WiMAX-internal Mobile IP layer. The considerations in this section are concerned with the outermost such layer, where the related Home Agent 3GPP HA resides in the 3G network. It is still to be decided if the HA is located in the SAE anchor, cf. architecture in section 4.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.1 General requirement
|
Major security threats related to IP mobility, when the procedures are not properly secured, are:
- IP address ownership needs to be verified else redirection attacks will happen
- Traffic sent to a target redirected elsewhere
- Attacker can blackhole traffic to a victim
- Attacker can insert itself on-path as a Man-in-the-Middle
- Redirecting traffic for someone to a victim
- Leads to (D)DoS (distributed denial of service) 3rd party bombing
- Consequently charging can be confused
- (D)Dos attack on mobility anchor
Key handling principle for inter-3GPP HO:
Before handover from EUTRAN to non-3GPP IP access network and/or from non-3GPP IP access network to EUTRAN, UE and EPS core network use the present key and the same key derivation function to derive the new key, which is to be used after handover.
(From S3-070732)
Some of the main problems that need to be considered when defining secuity context transfer optimizations for non-3GPP/3GPP handovers are:
• Security (avoiding negative impact on LTE/UMTS security)
• User privacy related to identity management
• AAA architecure misaligmnent between 3GPP and non-3GPP accesses
• Difficulty of defining a unique reference point for (secure) inter-access security context transfer.
• Possible standardization impact outside 3GPP (IETF, IEEE).
These shall be taken into account when looking at optimizations for handovers between 3GPP and non-3GPP accesses.
There are different kinds of make-before-break solutions using pre-authentication. This pre-authentication could take place either at the time of hand-over preparation, or (for e.g. single-radio terminals) the authentication could (perhaps) be prepared at the initial attach. It is an agreed working assumption that solutions based on pre-authentication should be the focus of the SA3 study for authentication optimizations for handovers between 3GPP and non-3GPP accesses
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.2 Host based Mobility
| |
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.2.1 Security associations used with Mobile IP
|
Figure 1 gives an overview of the MIP security associations which need to be present irrespective of the version of Mobile IP used. More security associations may be required for certain versions of Mobile IP. E.g. for Mobile IP v4 with a Foreign Agent, security associations between MN and FA, and FA and HA are needed.
Figure 1: Overview of the security architecture for MIP
The needed security associations are:
- A security association between the UE and 3GPP AAA. It is assumed that the 3GPP AAA in HPLMN is in charge of user authentication and authorization. This security association is based on a long-term secret.
- A security association between the UE and 3GPP MIP HA. This security association is established dynamically.
- A security association between 3GPP MIP HA and 3GPP AAA server in the same network. Typically, this security association is static. NDS/IP could be used when proxy AAA is used in roaming case. See TS 33.210 for more detail information [4].
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.2.2 Security protocols used with Mobile IP
|
1. The security association between the MN and 3GPP AAA is used for (mutual) authentication. In our context, the authentication protocol may be e.g. EAP-AKA. This protocol is independent of Mobile IP, but keys derived from a run of this protocol may be used for Mobile IP purposes.
2. The security association between the MN and 3GPP MIP HA is used for MIP signalling integrity protection. The protocols used depend on the version of Mobile IP. To give examples:
MIPv4: Home agent and mobile nodes shall be able to perform message authentication according to RFC 3344. MN-HA key agreed between HA and MN during MIP authentication is used to compute the digest in the Mobile-Home Authentication Extension according to RFC3344. The Mobile-Home Authentication Extension is used to provide integrity of signalling between Mobile Node and Home Agent. HMAC-MD5 shall be used as authentication algorithm with a key size 128 bit. HA will compute the UDP payload (RRQ or RRP data), all prior extensions, the type, length and SPI of the extension with MN-HA key in MIP Req-resp. MN uses with HMAC-MD5 to verify the received message from HA.
For MIPv4 with a foreign agent, more security associations are needed, as mentioned in the previous subsection. RFC3344 can also be used for these. The foreign agent shall be able to support message authentication using HMAC-MD5 and key size of 128 bits, with a key distribution mechanism (FFS).
MIPv6: IPsec is specified as the means of securing signalling messages between the Mobile Node and Home Agent for Mobile IPv6 (MIPv6) in RFC3776. RFC4285 proposes an alternate method for securing MIPv6 signalling messages between Mobile Nodes and Home Agents. The alternate method consists of a MIPv6-specific mobility message authentication option that can be added to MIPv6 signalling messages.
The alternate method is entirely based on shared secrets and does not use IPsec.
3. The security association between 3GPP MIP HA and 3GPP AAA server in the same network is used to securely transport the MN-HA keys from AAA server to MIP HA. It may not be needed if the interface between AAA server and HA is secured by other means.
Home agent and mobile nodes may perform message authentication whenever it is needed.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.3 Bootstrapping of Mobile IP parameters
| |
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.3.1 General
|
It would be undesirable for SAE if the UE had to obtain security credentials to be used specifically for Mobile IP signalling security. Rather, the security associations required for Mobile IP should be able to be derived from security credentials already available. In the case of SAE, this means that it should be possible to derive the security associations required for Mobile IP from the USIM.
Authentication between the MN and the network shall be performed as. A subscriber, who wants to use MIP, will have its subscriber profile located in the 3GPP AAA in the Home Network. The subscriber profile will contain information on the subscriber that may not be revealed to an external partner, At MIP registration , during a change of location between different access networks by matching the request with the subscriber profile, if the subscriber is allowed to continue with the request or not.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.3.2 RFC3957 used in conjunction with GBA
|
NOTE: this subsection applies only to MIPv4.
MN-HA key generation & distribution based on RFC 3957. This method uses pre-shared secret between MS and AAA server to establish a shared secret between MS and HA and / or MS and FA.
Figure 2: MN-HA key generation & distribution
1. During initial MIPv4 registration, MS includes a new extension (called the MN-HA Key Generation Nonce Request extension [RFC 3957]) in RRQ to request for a nonce from HAAA. The RRQ also contains the MS’s credential in the MN-AAA authenticator extension.
2. FA sends DIAMETER/RADIUS Access-Request to HAAA to authenticate the MS credential.
3. If the MS is authenticated successfully, the HAAA returns DIAMETER/RADIUS Access-Accept.
4. FA forwards the RRQ to the HA.
NOTE: If co-located care-of address mode is used, then RRQ message will be sent from MS to HA directly without FA in above picture
5. HA sends DIAMETER/RADIUS Access-Request to HAAA. In case of Roaming, the message will send through VAAA to HAAA. The DIAMETER/RADIUS Access-Request contains the MN-HA SPI attribute to request for a MN-HA key to HAAA that the MN-HA key needs to be derived. The HA may include the MS credential in the DIAMETER/RADIUS Access-Request.
Editor’s note: it’s FFS if it’s possible for a HA in the visited network.
6. HAAA selects a nonce and derives the MN-HA key from the MN-AAA shared secret, MS’s NAI, and the nonce.
7. HAAA returns DIAMETER/RADIUS Access-Accept that contains the MN-HA key and the nonce.
8. The HA sends RRP with a new extension (called the Generalized MN-HA Key Generation Nonce Reply Extension [RFC 3957]) carrying the key generation nonce, and the MN-HA authenticator computed from the MN-HA key. The new extension must precede the MN-HA authenticator. (FA forwards the RRP to the MS)
9. The MS derives the MN-HA key and uses it to verify the MN-HA authenticator in the RRP.
One possible way is to use GBA in conjunction with RFC 3957. In this case HAAA is associated with NAF.
Figure 3: Using GBA to derive and distribute MN-HA Keys (HAAA as NAF)
Generic Bootstrapping Architecture (GBA) allows bootstrapping of shared secrets between a UE/MN and the home network (Bootstrapping Service Function, BSF), which can then be used to derive further shared secrets to be used between MS and a Network Application Function(NAF).
Two options for using GBA in the inter access mobility authentication are considered:
- using GBA to derive the MN-HA Keys, in which case the HA is used as NAF and.
- using GBA to provision MN-AAA Keys, in which case HAAA is used as a NAF.
Figure 5 shows how GBA could be used to derive and distribute MN-HA Keys when HAAA as NAF, i.e. HAAA is associated with a Network Application Function (NAF).
1. The MN performs a bootstrapping procedure with the BSF and generates a (master) shared secret, Ks. Bootstrapping procedure is performed between the UE/MS and the BSF (which is located in the home network). During bootstrapping, mutual authentication is performed between the MS and the home network, and a bootstrapping key, Ks, will be generated by both the UE/MS and the BSF. Associated with the Ks include a Bootstrapping Transaction Identifier (B-TID) and a lifetime of the Ks.
NOTE: This procedure is only needed during initial registration (and it can be done before the MIP registration). It is not repeated at every HO (Handover). The only time it needs to be repeated is when the key is about to expire. But even in this case, the GAA procedure is done “offline”—i.e. the next MIP registration does not need to wait for GAA procedure to complete.
2. MN can then start MIP related signalling with the HA, which in turn contacts the HAAA.
3. HA then contacts to HAAA using Diameter/ RADIUS. Note: in the baseline document only RADIUS message is shown in the figure and the text. However, both Diameter and RADIUS can be used.
4. The HAAA, acting as a NAF, does not have the MN-AAA key, as the MN-AAA key is supposed to be generated by the BSF using Ks and other inputs to a KDF (key derivation function). Therefore, the HAAA will contact the BSF and fetch the MN-AAA key (Ks_(ext/int)_NAF of the HAAA) needed to authenticate the MN.
5. MN-HA keys are then derived from the MN-AAA Key using RFC 3957.
NOTE: If foreign agents (FA) are used, then foreign agent use Diameter/RADIUS to communication with HAAA.
Editor’s note: it needs to check how to send the B-TID in MIP registration message.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.3.3 Use GBA to generate MN-HA key
|
NOTE: This subsection applies to MIPv4and MIPv6.
In this alternative authentication method, HA is associated with NAF.
Home Agent (HA) is associated with a NAF, and Ks_(ext/int)_NAF would be used as MN-HA key: the MN performs a bootstrapping procedure with the BSF and generates a (master) shared secret, Ks. After that, the MN can start MIP related signalling with the HA, which in turn contacts the BSF to fetch MN-HA key.
Figure 4: Overview of GBA operations
1. Bootstrapping procedure is performed between the UE/MS and the BSF (which is located in the home network). During bootstrapping, mutual authentication is performed between the MS and the home network, and a bootstrapping key, Ks, will be generated by both the UE/MS and the BSF. Associated with the Ks include a Bootstrapping Transaction Identifier (B-TID) and a lifetime of the Ks.
NOTE: This procedure is only needed during initial registration (and it can be done before the MIP registration). It is not repeated at every HO (Handover). The only time it needs to be repeated is when the key is about to expire. But even in this case, the GAA procedure is done “offline”—i.e. the next MIP registration does not need to wait for GAA procedure to complete.
2. Once bootstrapping is completed, UE/MS can make use of the bootstrapped security association with a network application server, called the Network Application Function (NAF). To do so, the UE/MS communicates with the NAF. The UE/MS conveys to the NAF the B-TID.
3. The UE/MS derives the application specific session keys Ks_(ext/int)_NAF using a pre-defined key derivation function (KDF), with Ks, identifier of the NAF (NAF_Id), as well as other information as input. Upon receiving the request from UE/MS in step 2, the NAF contacts the BSF over the Zn to request the Ks_(ext/int)_NAF. The NAF provides the B-TID received from the UE/MS, and provides its own identity (NAF_Id). The BSF derives the Ks_(ext/int)_NAF in the same way as the UE/MS, and returns the derived key to the NAF. The Ks_(ext/int)_NAF can then be used as the shared secret between the MS and the NAF for any further security operations.
NOTE: If foreign agents (FA) are used, then foreign agent implements GAA NAF to get the MN-FA key.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.3.4 Use partial GBA to derive MN-HA Keys
|
NOTE: This subsection applies to MIPv4 and MIPv6.
GBA was designed for a situation where a UE wants to securely access potentially many application servers (NAFs), while having to be authenticated to the home network (and consume authentication vectors) in the Ub protocol run only once. Furthermore, the NAFs the UE wants to access may and need not be known at the time of the Ub protocol run. These requirements do not apply to MIP bootstrapping: the number of MIP servers with which the UE needs to share a key is limited to one, namely the Home AAA or Home Agent (when no Foreign Agent is used), and two, when an FA is used (or three, when two FAs are involved in a handover situation). In addition, the addresses of HA and FA cannot be chosen by the UE any time later, but are assigned by the home network (HA) and the visited network (FA), respectively. Therefore, the full functionality of GBA may not be needed.
A disadvantage of the use of GBA for MIP bootstrapping is that the HA, and, if applicable, the FA, need to support NAF functionality. An off-the-shelf HA or FA does not do that.
Editor’s note: the intention of this GBA extension is a subset of GBA and should not be a problem.
We consider two cases below. For both cases, the following is assumed:
- a UE has to run the Ub protocol with the BSF before starting MIP registration.
- the BSF is integrated with the AAA server (as in the current baseline document).
- the AAA server distributes keys to HA and FA using standard AAA procedures (for MIPv4: RFC4004: DIAMETER Mobile IPv4 application, and for MIPv6: draft-ietf-dime-mip6-split-03), and does not use the Zn interface.
- the distributed keys are used with the Mobile IPv4 and Mobile IPv6 authentication mechanisms defined in RFC 3344 and RFC 4285 respectively
Editor’s note: it’s FFS whether RADIUS extension also needs to be supported.
With these assumptions, HA and FA can be off-the-shelf, and need not be GBA-aware. The Ua and the Zn interfaces are not needed.
Case 1: HA and FA addresses and/or names are acquired by the UE independently of the Ub protocol run
In this case, the BSF and the UE derive keys Ks_(ext/int)_NAF to be shared between UE and HA, and UE and FA, respectively, as specified in TS 33.220.
Editor’s note: no change to Ub in Case 1.
Case 2: The HA address and/or name is acquired by the UE as part of the Ub protocol run
In this case, the BSF can send the FQDN, and possibly also the IP address, of the HA to the UE in a new element in the XML body of the “200OK” message, which is the last message in the Ub protocol run. This provides an alternative to SAE HA address assignment. Note that it may not be obvious for all access systems how to let the UE acquire the SAE HA address.
Editor’s note: the Ub interface will be affected in Case 2.
The FA address needs to be acquired by the UE locally.
The use of partial GBA for MIP bootstrapping is captured in Figure 5.
Figure 5: Partial GBA for MIP bootstrapping
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.3.5 Using IKEv2
|
Authentication between the MN and the network and IPsec SA setup between the MN and the HA for MIPv6 shall be performed using IKEv2 as defined in the IETF draft [draft-ietf-mip6-bootstrapping-split-02.txt]. In SAE, the home agent communicates with the AAA server to perform mutual authentication. The IKEv2 authentication is performed using EAP-AKA.
Figure 6: MN-Network authentication and MN-HA IPsec SA setup for MIPv6
Editor’s note 1: adding relatively heavy protocol of IKEv2 should be considered to be for further study if cost efficiency is in appropriate level.
Editor’s note 2: this is only one of multiple different options.
Editor’s note 3: both I-WLAN scenarios 2 and 3 should be studied
(From S3-070820)
The first procedure that must be performed by the MN is the discovery of the HA address, which in case of EPS is the IP address of the PDN GW.
As soon as the Mobile Node has discovered the PDN GW address, it establishes an IPsec Security Association with the Home Agent itself through IKEv2. The detailed description of this procedure is provided in RFC4877. The IKEv2 Mobile Node to Home Agent authentication is performed using Extensible Authentication Protocol (EAP).
When the Mobile Node runs IKEv2 with its Home Agent, it shall request an IPv6 Home Address through the Configuration Payload in the IKE_AUTH exchange by including an INTERNAL_IP6_ADDRESS attribute. When the Home Agent processes the message, it allocates a HoA and sends it a CFG_REPLY message. The IPv6 Home Address allocation through IKEv2 allows to bind the Home Address with the IPsec security association so that the MN can only send Binding Updates for its own Home Address and not for other MN’s Home Addresses.
Figure 7 provides the flow for the initial DS-MIPv6 bootstrapping.
Figure 7: DS-MIPv6 bootstrapping based on IKEv2
1) The UE discovers the PDN GW address based on the procedure specified in 23.401.
2) The UE starts an IKEv2 exchange with the PDN GW. The first part of this exchange is an IKE_SA_INIT exchange.
3) The UE indicates that EAP is used for IKEv2 authentication and an EAP exchange is performed. EAP is carried over IKEv2 between the UE and the PDN GW and over the AAA protocol between the PDN GW and the AAA server.
4) During the IKEv2 exchange, the PDN GW allocates an IPv6 Home Address and send it to the UE in a IKEv2 Configuration Payload.
5) As a result of the previous steps, an IPsec SA is established to protect DS-MIPv6 signalling.
6) The UE sends the MIP Binding Update message to the PDN GW.
7) The PDN GW processes the binding update. The PDN GW sends the MIP Binding Ack to the UE.
8) As a result of the above steps a MIPv6 tunnel is established and the UE can start using its home address at the application level.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.3.6 Security bootstrapping for DS MIPv6 using MIP options
|
(From S3-070748)
This procedure uses the MIP authentication options defined in RFC4285 [8] to provide authentication of Binding Update and Binding Acknowledgement messages, namely the
• MN-HA Mobility Message Authentication Option and the
• MN-AAA Mobility Message Authentication Option.
The AAA Mobility Message Authentication Option is used when the MN and the HA do not yet have a shared key, i.e. in the situation requiring bootstrapping of the MN-HA key.
It is assumed that the MN and the AAA server share a long-lived security association.
NOTE: It is ffs whether there is a need to dynamically generate the MN-AAA key and, if so, how to do it. Alternatives would include derivation during network access authentication and GBA.
The MN-HA key is derived from the MN-AAA key and a nonce. The nonce is requested by the MN in a Key Generation Nonce Request option and provided by the AAA server to the MN in a Key Generation Nonce Reply option. These options are described in draft-devarapalli-mip6-authprotocol-bootstrap [9].
NOTE: Instead of using a nonce for generating the MN-HA key from the MN-AAA key, also the timestamp from the Mobility Message Replay Protection Option, cf. below, could be used. This is ffs.
The HA may provide a Home Address to the MN using the Home Address Options defined in draft-devarapalli-mip6-authprotocol-bootstrap [9].
The communication between the Home Agent and the AAA server is based on DIAMETER extensions described in draft-ietf-dime-mip6-split [10]. This communication is assumed to be authenticated (integrity-protected).
Figure 8: Bootstrapping using Mobile IP options
Description of the information flow in Figure 8:
1. When the Mobile Node (MN) does not yet share a key with the Home Agent (HA) the MN sends a DSMIPv6 Binding Update (BU) including the MN-AAA authentication mobility option. The MN also includes a Key Generation Nonce Request Option. If the MN does not yet have a Home Address (HoA) it also includes the Home Address Request Option in the BU. The MN shall include the Mobility Message Replay Protection Option defined in RFC 4285 [8] containing a timestamp.
2. When the Home Agent receives a BU with the MN-AAA mobility message authentication option, the HA forwards the BU to the AAA server for authentication.
3. The AAA server authenticates the BU by verifying the message authentication code in the MN-AAA authentication mobility option, using the MN-AAA shared key and the timestamp in the Mobility Message Replay Protection Option.
4. Upon successful authentication of the BU, the AAA server sends the parameters of the MN-HA security association (key, algorithm) to the HA. The AAA server also returns a nonce and algorithm identifier in the Key Generation Nonce Reply Option.
5. The HA sends a Binding Acknowledge (BA) message protected with the MN-HA security association received from the AAA server to the MN. The HA forwards the Key Generation Nonce Reply Option as part of the BA. The HA also includes the Assigned Home Address Option in the BU if the MN requested a HoA. The HA checks the validity of the timestamp and, if necessary, includes an indication of a timestamp mismatch, as described in RFC 4285 [8]. In the latter case, HA deletes the MN-HA security association after sending the BA.
6. The MN generates the MN-HA key from the MN-AAA key and the nonce. The MN then verifies the BA using the MN-HA authentication mobility option. If the BA contains an indication of a timestamp mismatch the MN resends the BU from step 1, but with the message authentication code in the MN-AAA authentication mobility option computed over the corrected timestamp.
7. For subsequent BUs, the MN uses the established MN-HA security association and does not include an MN-AAA authentication mobility option.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.4 Network based Mobility
| |
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.4.1 PMIP
| |
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.4.1.1 Introduction
|
This section looks at how PMIP messages need to be protected within the Evolved Packet Core and how PMIP protection needs to be handled if the PMIP messages originate from a trusted non-3GPP network node.
This analysis is based on draft-ietf-netlmm-proxymip6-06.txt [11] from which in particular the sections 4 and 11 have been used from a security viewpoint.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.4.1.2 Overview of PMIP usage in 3GPP
|
(From S3-070756)
PMIPv6 defines a MAG (Mobile Access Gateway) and an LMA (Local Mobility Anchor) from which the LMA will be integrated in the PDN Gateway or Serving Gateway (for the roaming case).
Figure 9: Protocols for MM control and user planes of S2a for the PMIPv6 option
TS 23.402v130 section 5 is relevant in this respect and specifies that PMIPv6 may be used on following reference points:
• S2a: Between a node in the trusted non-3GPP access network (Foreign agent) and the LMA (Home Agent)
• S2b: Between the ePDG and the LMA (Home Agent).
TS 23.402v130 section 4.2.1 mentions the use of PMIP based S5 reference point between the Serving Gateway and the PDN Gateway. The S5 reference point may also apply GTP, and is an intra-operator interface. PMIP usage over S5/S8b is currently included in the description of PMIP use over S2b, and (see section 5.4.2.4.3 TS 23.402) in case of roaming, the S-GW is the LMA for PMIP procedure in S2b between the ePDG and the S-GW and the PDN GW is the LMA for PMIP procedure in S8b between the S-GW and the PDN GW. In addition, PMIP over S5/S8b is discussed in section 5.4.2.6 TS 23.402 for E-UTRAN access.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.4.1.3 PMIP trust model
|
PMIPv6 is an IETF based network-based mobility management mechanism, and has applied the same trust model properties as the use of GTP for mobility management in UMTS and the EPC (for the S5 and S2b reference points). This means the MAG i.e. the Serving Gateway (S5) or ePDG (S2b), is sufficiently trusted by the LMA to register only those Mobile Nodes that are attached.
However when the MAG is located in a trusted non-3GPP network (S2a), there is a little bit of a difference to the current 3GPP or PMIPv6 draft [11] trust model where a 3GPP network component (SGSN, S-GW) is trusted to register only attached MNs. Here, the MAG could e.g. be located in a WLAN AP which can much more easily be tampered with than an SGSN or S-GW. The implication of this scenario is for ffs (see also proposed decision at the end of this section).
The trust between the LMA and the MAG is verified by the LMA by allowing only those MAGs to perform Binding Updates which are known by the LMA i.e. by the use of IKEv2 authentication. This measure defends against a Network Node trying to impersonate another MAG, and thus will protect against Denial-Of-Service attacks from the Mobile Node's viewpoint.
The PMIPv6 draft [11] recognizes the threat of a compromised MAG that would send PMIP messages on behalf of a Mobile Node with a Mobile Node not present on the local link. From section 11 of [PMIPv6 draft]:
"To eliminate the threats related to a compromised mobile access gateway, this specification recommends that the local mobility anchor before accepting a Proxy Binding Update message for a given mobile node, should ensure the mobile node is definitively attached to the mobile access gateway that sent the binding registration request.
The issues related to a compromised mobile access gateway in the scenario where the local mobility anchor and the mobile access gateway in different domains, is outside the scope of this document. This scenario is beyond the applicability of this document."
The last sentence from the extract is an indication for the fact that the S2a use is not covered by PMIPv6 draft [11] and needs additional considerations.
Although required by PMIPv6 draft [11] it is unclear how the LMA should be able to verify that the MN has attached, rather this seems to be a property of the PMIP model that the MAG is trusted to apply those requirements. The authorization mechanisms on the MAG-LMA interfaces are inadequate for this.
The effect of a potential misuse by the MAG could be limited to those MAGs on which the Mobile Node is authorized to attach. This authorization shall then be verified by the LMA. However, this explicit authorization-check may be cumbersome to administrate per user (and therefore not very effective), and if not administrated per user but per roaming partner, the authorization check rather takes place between the MAG and the LMA (via the lack of shared secrets for IKEv2, or certificate authorization checks), and this fits the PMIP trust model applying to S5 and S2b.
Extending PMIPv6 by involving the UE in order to produce a fresh user involvement on the MAG that can be used towards the LMA, is a contradiction to the design guidelines of PMIPv6: "This protocol enables mobility support to a host without requiring its participation in any mobility related signaling." Furthermore verifying the user involvement would also increase the amount of signaling needed. So there is a trade-off between trust/security and amount of signaling.
NOTE 1: For the other network based mobility management protocols e.g. GTP this has worked well in the past. The operator should be able to trace down suspicious registrations as long as the links are secured (physical or by NDS/IP).
In case of S5, the node implementing the MAG may already be trusted to receive an EPS security context for a user, without proof of user involvement.
NOTE 2: In case S-GW and MME are implemented on the same physical node.
The risk caused by a misuse of the received key material is greater than the risk due the use of the PMIPv6 trust model. Verifying user involvement during mobility management registration would need to involve an additional authentication verifiable by the LMA only such that the compromised MAG cannot impersonate the user, where then we are back to the DSMIPv6 solution.
Conclusion:
a) use PMIPv6 as defined by IETF [draft-ietf-netlmm-proxymip6-06.txt] for S5 and S2b
b) if the trust relation between the MAG and the LMA is not there then additional security measures are needed. These security measures are for ffs.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.4.1.4 Security measures on the Reference points between the LMA and the MAG that have a trust relation
|
PMIPv6 draft [11] section 4 recommends the use of IPsec ESP in Transport Mode (RFC4303) as default security mechanism for integrity protection and data origin authentication for PMIP messages and IKEv2 end-to-end between the MAG and the LMA to establish IPsec security associations. Confidentiality protection of PMIP messages is not required.
Section 5.5.1 allows the use of one security tunnel between the MAG and the LMA instead of a dynamic set-up.
"The bi-directional tunnel is established after accepting the ProxyBinding Update request message. The created tunnel may be shared with other mobile nodes attached to the same mobile access gateway and with the local mobility anchor having a Binding Cache entryfor those mobile nodes. Implementations MAY choose to use static tunnels instead of dynamically creating and tearing them down on a need basis."
Therefore alternatives to the IKEv2 usage like NDS/IP (TS 33.210) should still be possible (RFC 2406 and IKEv1) and can provide the same security services.
The only difference is the hop-by-hop approach with SEGs (requiring tunnel mode towards the SEG), which should not be a problem in viewpoint of security if the network owning the SEG and the LMA is sufficiently trusted. The use of TS 33.310 is needed when LMA and MAG belong to a different operator.
The PDN gateway may already implement IKEv1/IPsec for protecting the signaling towards the AAA/HSS in case of DSMIPv6 and may already implement IKEv2 in case that such mechanism would be selected for DSMIPv6 protection towards the Mobile Node (which is for ffs at SA3#49). The ePDG already requires IKEv2 implementation towards the UE.
Conclusion:
SA3#49 agreed that the choice between IKEv1 (as defined by NDS/IP) or IKEv2 (as proposed by PMIPv2] for PMIP message protection between the MAG and the LMA needs further study
a) Both IKEv2 and IKEv1 can provide the necessary security features.
b) Referring to NDS/IP (TS 33.210) and NDS/AF (TS 33.310) allows a hop-by-hop security model.
c) The difference between RFC2406 [which is referred by NDS/IP] and RFC4303 [which is referred by PMIPV6] is not essential for the decision.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.4.1.5 The need for using strong access authentication with Proxy Mobile IP
|
Clause 7.4.1.3 discusses the need for trust of the LMA in the correct operation of the MAG. Trust in the MAG means that the LMA can be ensured that the operation of MAG is not somehow influenced by an attacker. Clause 7.4.1.4 discusses the security on the reference point between the MAG and the LMA. Security on this reference point ensures that PMIP messages are originating from a trusted entity, and that no attacker could tamper with them in transit. This subclause discusses an additional requirement for the secure operation of PMIP: strong access authentication. In the context of PMIP, the authentication scheme shall be considered sufficiently strong by all stakeholders involved, in particular by the operators of MAG and LMA. In EPS the LMA is the PDN GW owned by a 3GPP EPS operator. This implies that the authentication scheme shall satisfy also 3GPP security requirements, i.e. it shall use a USIM.
PMIP is based on the assumption that a MAG can securely identify which user is attached to the access network served by the MAG. This secure identification is realised by access authentication. If access authentication was weak then an attacker could impersonate a user in the access network. If this happened, a MAG would report in good faith to the LMA that a certain user was present in the access network, while in fact the attacker was present. This could result in Denial of Service to the impersonated user through the use of PMIP because all traffic destined to this user would then be routed to a wrong destination.
An impersonation attack exploiting a weakness in access authentication could occur by attacking any part of the access network. Neither the trusted operation of the MAG nor the security on the reference point between the MAG and the LMA would prevent such an attack if access authentication was weak. In this sense, the requirement of using strong access authentication with PMIP is complementary to the requirements addressed in clauses 7.4.1.3 and 7.4.1.4.
Section 4 of TR 33.922 requires USIM-based authentication also for non-3GPP access. Currently, AKA is the only authentication scheme known to use the USIM. As AKA-based authentication is considered sufficiently strong, also the requirement introduced in this subclause is considered fulfilled in EPS.
Conclusion:
When PMIP is used within EPS, strong access authentication is required. In EPS and as per clause 4 of TR 33.922, this requirement is fulfilled since the USIM-based authentication for non-3GPP access is mandated. The USIM-based authentication implies the use of the AKA protocol, which is considered sufficiently strong.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.4.1.6 No trust relation between LMA and MAG on S2a
|
NOTE：This section describes the case when there is no trust relation between the MAG and the LMA. However, what this section describes is not aligned with the assumption of TS33.402v100 section 9.3.1.2.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.4.1.6.1 Security risks
|
MAG lies in the trusted non-3gpp IP access system in S2a. There may be no trust relation between the MAG and the LMA since they may belong to different operators. In this case, a compromised MAG may make an attack to UE in other MAG’s domain. Also, a compromised MAG may send fake PBU message to update the binding of UE who is served by other MAG. In this case,
-the victim UE cannot receive the data since the data is routed to the compromised MAG;
-the compromised MAG can eavesdrop the data of UE who is served by other MAG;
-the compromised MAG may send a large amount of PBU to make the LMA in burden and a DoS attack may occur;
PMIPv6 [draft-ietf-netlmm-proxymip6-11] defines to use IPsec to protect PBU/PBA. However, the prerequisite is that there should be trust relation between the MAG and the LMA. PMIPv6 security mechanism cannot work for the condition of no trust relation between the MAG and the LMA.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.4.1.6.2 Possible measures
|
One possible way is to have the mapping between the UE and the serving MAG in one of network servers. When a MAG sends a PBU to a LMA, the LMA can ask this server to check whether this MAG is currently serving the UE. In this way, it will be avoided that a compromised MAG represents UE served by other MAGs to send the fake PBU.
UE should run an EAP-AKA with MAG before PMIPv6 procedure. The AAA server in UE’s home network can record which MAG executed EAP-AKA procedure. In the meanwhile, AAA can keep the mapping between UE and its serving MAG. In this way, when a MAG sends a PBU message to a LMA, the LMA can ask the AAA server to have a check whether this MAG is serving the UE in the current time according to the identity of UE in PBU message. When UE moves to other MAG, AAA should know the change since AAA will be involved in changing MAG’s procedure. So the AAA can update the mapping between UE and the serving MAG.
Editor’s Note: Another solution is that AAA sends a key to MAG which is related to the UE after EAP-AKA procedure. UE will participate in the EAP-AKA. So the MAG can obtain this key only when this MAG really serves the UE. This key can be used to protect the integrity of PBU messages. LMA interacts with AAA to check the integrity of PBU message. In this way, a compromised MAG cannot get the related key. So it can not send valid PBU message. When UE moves to other MAG, AAA should know the change since AAA will be involved in changing MAG’s procedure. So the AAA can update the related key and send the key to the current serving MAG. This solution may need clarify and FFS.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
7.4.2 NetLMM
|
In Network based Localized Mobility Management (NetLMM) the Localized Mobility Anchor (LMA) is configured with a globally routable network prefix which the IP address assigned to UE is composed of, and packets to/from the UE are tunnelled between LMA and Mobile Access Gateways (MAGs). MAG shares the same network prefix as LMA’s one, therefore, when the MN moves from one MAG to another, neither the subnet nor the MN IP address are changing. Here the LMA handles the packets incoming from Internet to the operator’s domain. Each MAG is configured with the information needed to contact the LMA. This is also depicted in Figure 10.
Figure 10: Proxy mobility protocol scenario
NetLMM does not bring any additional security threats. The protocol does face the general security threat of IP address ownership that is valid for all mobility protocols. Solution for this threat is to:
- Secure Link Layer attachment (Packet Data Protocol Context secured by 3G AKA)
- IP address allocation by the network over secured attachment.
For NetLMM the countermeasure regarding the security threats in Section 7.1 are:
- IP address ownership
- Enforce IP address ownership at network attachment. IP address is allocated by network (e.g., DHCP, PDP) over secure network attachment (e.g., 3G AKA). IP address binding is enforced during communication.
- (D)DoS attack
- Attack on forwarding resources
- Requires knowledge of the network prefix allocated for MNs
- Outside Correspondent Node and MNs are aware
- Attack on control plane endpoint resources
- Requires knowledge of the anchor point IP address
- NetLMM LMA IP address is hidden from MNs and outside CNs.
- NetLMM shall be resilient to DoS because only the forwarding resources can be attacked. Those can be dealt with by over-provisioning the forwarding capacity.
Editor’s note: further details should be added.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
8 Specific aspects of security for mobility between 3GPP access systems and non-3GPP access systems
| |
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
8.1 Security for mobility between pre-SAE 3GPP access systems and non-3GPP access systems
|
It needs to be clarified in the course of the work on SAE mobility to what extent mobility, and, in particular the related security aspects involving pre-SAE 3GPP access systems require a different handling. The goal is, of course, to minimise or completely avoid the differences, but it is currently not clear in how far this goal can be achieved.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
8.2 Security context transfer between 3GPP and trusted non-3GPP access networks
|
Security context is the information on the current state of a UE in the serving system required to re-establish the security association in the target system. Security context includes
1. Agreed security algorithms between the UE and the serving network,
2. Agreed/derived encryption and/or integrity protection keys and key identifiers.
3. Security association related information like key lifetime, sequence number, count values etc.
4. The temporary identity issued by the serving network
NOTE: In 3GPP, temporary identity is used by the target network to identify the serving network, but it’s FFS for handover between 3GPP and non-3GPP networks whether temp IDs to be used for identifying the pervious access network.
As 3GPP has already adopted security context transfer procedures for optimizing authentication during handover, it is reasonable for SAE to enable security context transfer between the 3GPP and non-3GPP networks.
Editor’s note: the content of security context is FFS.
|
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
8.3 ANDSF Security
| |
f1b0a2c921a1d71d732d8ab589640b1a
|
33.922
|
8.3.1 General
|
ANDSF (Access Network Discovery and Selection Function) is a mechanism of access network discovery and selection. It is provided in order to control the UE's inter-system handover decisions and in order to reduce the battery consumption for inter-system mobility. See TS23.402 for more details. However, the privacy of UE and the operator needs to be protected if private information will be sent between UE and ANDSF server. Reusing GBA and PSK TLS to establish SA between UE and ANDSF server will be easily implemented by both operators and vendors. PSK TLS can be used for the security association between UE and ANDSF server. It can provide confidentiality and integrity protection for ANDSF security.
8.3.2 Procedure
Figure 11: ANDSF security using GBA
1. The UE and the BSF will process bootstrapping procedure. The master key Ks will be derived in this procedure.
Editor’s note: It is FFS if 3GPP AAA can be the BSF in this scenario to be easily deployed by the operator.
2. The UE discovers the ANDSF server. See more details in TS23.402. Then the UE derives the Ks_ANDSF.
3. The UE starts communication with ANDSF server. UE sends application request to ANDSF server.
4. ANDSF server sends authentication request to the BSF for the key,
5. The BSF derives the Ks_ANDSF based the master key Ks. The derivation function is the same with Ks_NAF. And then BSF sends Ks_ANDSF and the key lifetime to the ANDSF server.
6. ANDSF server will inform UE that it gets the key Ks_ANDSF and can continue the ANDSF function,
7. The UE and the ANDSF server establish the security association based on the Ks_ANDSF. The detailed method i.e. PSK TLS, can be referenced to TS24.109.
Editor’s note: It is FFS which method can also be used to establish the security association between the UE and the ANDSF server.
8. The UE and ANDSF server runs handover with ANDSF procedure after the SA was successfully established to protect the communication between them.
Annex A:
RFC 3957
From RFC 3957: “When the mobile node shares an AAA Security Association with its home AAA server, however, it is possible to use that AAA Security Association to create derived Mobility Security Associations between the mobile node and its home agent, and again between the mobile node and the foreign agent currently offering connectivity to the mobile node. …[RFC3957] specifies extensions to Mobile IP registration messages that can be used to create Mobility Security Associations between the mobile node and its home agent, and/or between the mobile node and a foreign agent.”
Appendix B of RFC3957 contains message flows for Requesting and Receiving Key Generation Nonce:
MN FA AAA Infrastructure
<--- Advertisement-----
(if needed)
- RReq+AAA Key Req.-->
--- RReq + AAA Key Req.--->
<--- RRep + AAA Key Rep.---
<-- RRep+AAA Key Rep.--
Annex B:
Change history
Change history
Date
TSG #
TSG Doc.
CR
Rev
Subject/Comment
Old
New
2006-02
Creation of document
0.0.0
0.0.1
2006-07
Revision of the document
0.0.1
0.0.2
2006-11
Revision of the document
0.0.2
0.0.3
2007-05
Including 3.1of S3-070399
0.0.3
0.0.4
2007-07
Including 2.1of S3-070506, and S3-070531
0.0.4
0.0.5
2007-10
Including S3-070748 and S3-070820, S3-070732 and S3-070756
0.0.5
0.1.0
2007-12
Including S3a070980.
0.1.0
0.2.0
2008-02
Including S3-080049, S3-080129.
0.2.0
0.3.0
2008-03
Correct the release number.
0.3.0
0.3.1
2008-04
Including S3-080362.
0.3.0
0.4.0
2008-06
Including S3-080725, S3-080765.
0.4.0
0.5.0
2008
MCC clean up for presentation to SA
0.5.0
1.0.0
Technical Specification Group Services and System Aspects TSGS#41(08)0479
Meeting #41, 15 - 18 September 2008,
Kobe, Japan
Presentation of Specification to TSG
Presentation to: TSG SA Meeting #41
Document for presentation: TR 33.922, Version 1.0.0
Presented for: Information
Abstract of document:
TR 33.922 is currently in inconsistent shape. It was useful during a certain period of the work towards TS 33.402, but the material in TR 33.922 was not updated when decisions contradicting or superseding the text in the TR were taken. Nevertheless, TR 33.922 can serve a useful purpose similar to TR 33.821 in relation to TS 33.401 on E-UTRAN security, namely to document the discussion process in 3GPP SA3, which led to the final version of the TS.
In SA3 #52 meeting it was agreed to add a disclaimer to the “Scope” section of TR 33.922 to make it clear that text in the TR may be inaccurate or outdated.
Changes since last presentation:
This is the first time that the document is presented.
Outstanding Issues:
There are no outstanding issues.
Contentious Issues:
There are no contentious issues.
|
566267d8c54c8da6ca65b0a2ce396207
|
32.807
|
1 Scope
|
The purpose of this document is to contain the updated Work Item Descriptions (WIDs) and capture status of all TSG SA WG5 work items of the current 3GPP Release in order for the group to get an overview of current ongoing work.
This TR is used as a mean to provide input to the complete 3GPP work plan that is handled by MCC.
|
566267d8c54c8da6ca65b0a2ce396207
|
32.807
|
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] http://www.3gpp.org/ftp/Information/WORK_PLAN/
[2] http://www.3gpp.org/ftp/Information/WI_Sheet/
SA5 Work Plan snapshoot
Unique_ID
Name
Acronym
% Cpl
Impacted TSs and TRs
Rapporteur
35041
OAM&P
OAM7
28%
Christian TOCHE
35042
Network Infrastructure Management
OAM7-NIM
28%
35044
Enhance NRM to accommodate NGN (IMS as basis of the Next Generation Network)
OAM7-NIM-NGN
10%
32.632, 32.633, 32.634, 32.635
BT
35045
IRP usage scenarios
OAM7-NIM
0%
new TR 32.8xy
Lucent
35046
Co-operative Element Management interface (CO-OP)
OAM7-NIM-COOP
75%
new TR 32.8xy, 32.101
Motorola
35047
Network Management (NM) Itf-N performance criteria
OAM7-NIM
30%
new TS 32.yzx
China Mobile
35048
Delta synchronization between IRP Manager and IRP Agent
OAM7-NIM
55%
new TR 32.8xy,new TSs 32.111-n,32.60n; For n = 1 to 5
Siemens
35049
Subscription Management (SuM) IRP Solution Sets
OAM7-NIM-SuM
55%
32.101,32.175, new TSs 32.161,32.307,32.317,32.607,32.617,32.667
Ericsson
35050
Integration Reference Point (IRP) Security Management
OAM7-NIM
40%
New TSs 32.372,32.373,32.374; 32.111,32.30x,32.32x,32.33x,32.34x,32.35x,32.36x,32.41x,32.60x,32.61x,32.66x
Huawei
35051
Integration Reference Point (IRP) Methodology
OAM7-NIM
15%
New TSs 32.15w,32.15x,32.15y,32.15z; 32.150,32.151,32.152
Ericsson
35052
Partial suspension of Itf-N during maintenance/testing
OAM7-NIM
50%
new TR 32.8xy, new TSs 32.30n,32.60n,32.61n; For n = 1 to 5
Siemens
35053
Advanced Alarming on Itf-N
OAM7-NIM
35%
new TR 32.8xy ; New TSs 32.111-n,32.30n; For n = 1 to 5
Siemens
35054
Management of Legacy Equipment
OAM7-NIM
20%
new TS 32.xxx, New TSs 32.62n,32.63n,32.64n,32.65n,32.71n,32.74n,32.69n; For n = 1 to 5
Siemens
35055
Rules for Vendor Specific Extensions
OAM7-NIM
20%
32.62n; For n = 1 to 5
Siemens
35056
CN CS Bearer Transport Network (BTN) relative NRM
OAM7-NIM
45%
New Tss TS 32.xx1,TS 32.xx2,TS 32.xx3,TS 32.xx4
China Mobile
35064
Backward and Forward Compatibility of IRP systems
OAM7-NIM-BFC
10%
TR 32.805. new TS 32.15X
Ericsson
35065
Study of Element Operations Systems Function (EOSF) definition
OAM7-NIM- EOSF
15%
new TR 32.8xy
China Mobile
35066
Study of SOAP/HTTP IRP Solution Sets
OAM7-NIM-SOAP
20%
new TR 32.8xy
Nortel
35067
Study of Itf-N Implementation Conformance Statement (ICS) template
OAM7-NIM-ICS
15%
new TR 32.8xy
China Mobile
35068
Study of IRP Information Model
OAM7-NIM-IM
30%
new TR 32.8xy
Motorola
35071
Repeater Network Ressource Model (NRM) definition
OAM7-NIM
5%
32.64n; For n = 1 to 5
China Mobile
35072
UTRAN radio channel power monitoring
OAM7-NIM
20%
32.403, 32.64n; For n = 1 to 5
China Mobile
35074
NEW Study on SA5 MTOSI XML Harmonization
OAM7-NIM-XML
0%
new TR 32.8xy
Nortel
35043
Performance Management
OAM7-PM
33%
35057
Performance measurements definition for CN CS
OAM7-PM
65%
New TS 32.xyz
China Mobile
35058
Enhancement UTRAN performance measurements definition
OAM7-PM
20%
32.403
China Mobile
35059
Add TDD specific counters in Performance measurement
OAM7-PM
75%
32.403
CATT
35060
ATM bearer network Performance measurements
OAM7-PM
30%
32.403
ZTE
35061
IP bearer network Performance measurement definitions
OAM7-PM
20%
32.403
China Mobile
35069
Performance measurements definition for IMS
OAM7-PM-IMS
5%
new TS 32.xyz
China Mobile
35073
HSDPA performance measurements
OAM7-PM
5%
32.403
China Mobile
35039
Trace Management
OAM7-Trace
20%
35040
Trace Management for IMS
OAM7-Trace-IMS
0%
32.421,32.422,32.423
Nortel
35062
End-to-end Service Level tracing for IMS
OAM7-Trace-IMS
20%
32.101,32.421,32.422,32.423
Vodafone
35070
IRP for Subscriber and Equipment Trace Management
OAM7-Trace-IRP
10%
32.421,32.422,32.423. 3 new TS 32.4x1,2,3
Nokia
35063
Trace record content for UTRAN TDD
OAM7-Trace-TDD
75%
32.421,32.422,32.423
CATT
Feature: Operations, Administration, Maintenance & Provisioning - OAM&P (OAM7) Unique_ID: 35041
Building Block: Network Infrastructure Management (OAM7-NIM) Unique_ID: 35042
Technical Specification Group Services and System Aspects TSGS#28(05)0302
Meeting #28, Quebec, CANADA, 06-08 June 2005
Source: SA5 (Telecom Management)
Title: WID WT Enhance NRM to accommodate NGN (IMS as basis of the Next Generation Network)
Document for: Approval
Agenda Item: 7.5.3
3GPP TSG-SA5 (Telecom Management) S5-050280
Meeting #42, Montreal, CANADA, 09 - 13 May 2005
Work Item Description
Title:
Enhance NRM to accommodate NGN (IMS as basis of the Next Generation Network) Unique_ID: 35044
Acronym: OAM7-NIM-NGN
1 3GPP Work Area
X
Radio Access
X
Core Network
Services
2 Linked work items
OAM&P (Operations, Administration, Maintenance & Provisioning) (Feature: OAM7)
WI
Unique_ID
OAM7
35041
Network Infrastructure Management (BB: OAM7-NIM)
WI
Unique_ID
OAM7-NIM
35042
3 Justification
The IMS has been adopted as the basis of the Next Generation Network (NGN).
It is proposed to enhance the 3GPP NRM in TS 32.63x Configuration Management (CM); Core Network Resources Integration Reference Point (IRP) - to accommodate any additional requirements identified.
4 Objective
In liaison with other groups (e.g. ETSI TISPAN, TeleManagement Forum (TMF), ITU-T SG4, Multiservice Switching Forum (MSF) to enhance the Core Network Resource Model to support the requirements of NGN Release 1 and Voice over IP (VoIP).
5 Service Aspects
None
6 MMI-Aspects
None
7 Charging Aspects
None
8 Security Aspects
None
|
566267d8c54c8da6ca65b0a2ce396207
|
32.807
|
9 Impacts
|
Affects:
UICC apps
ME
AN
CN
Others
Yes
X
X
No
X
X
Don't know
X
|
566267d8c54c8da6ca65b0a2ce396207
|
32.807
|
10 Expected Output and Time scale (to be updated at each plenary)
|
New specifications
Spec No.
Title
Prime rsp. WG
2ndary rsp. WG(s)
Presented for information at plenary#
Approved at plenary#
Comments
32.4xy
Subscriber and equipment trace; Trace Management Integration Reference Point (IRP): Requirements
SA5
3GPPSA#31
13 - 15 Mar 200625 - 27 Sep 2006
3GPPSA#34
Dec 2006
32.4xy
Subscriber and equipment trace; Trace Management IRP Information Service
SA5
3GPPSA#33
25 - 27 Sep 2006
3GPPSA#34
Dec 2006
32.4xy
Subscriber and equipment trace; Trace Management Integration Reference Point (IRP): CORBA Solution Set
SA5
3GPPSA#33
25 - 27 Sep 2006
3GPPSA#34
Dec 2006
Affected existing specifications
Spec No.
CR
Subject
Approved at plenary#
Comments
32.421
3GPPSA#34
4 - 6 Dec 2006
32.422
3GPPSA#34
4 - 6 Dec 2006
32.423
3GPPSA#34
4 - 6 Dec 2006
Reason for re-scheduling:
Recently SA5 identified some interworking between the Trace IRP and the Service Level Tracing (SLT).
For SLT, SA5 just agreed on the Requirements and the Trace IRP got some input to the IRP requirement.
11 Work item rapporteur(s)
[email protected] Toche
12 Work item leadership
SA5
13 Supporting Companies
Nortel, Nokia, Lucent Technologies, Huawei, Ericsson
14 Classification of the WI (if known)
Feature (go to 14a)
Building Block (go to 14b)
X
Work Task (go to 14c)
|
566267d8c54c8da6ca65b0a2ce396207
|
32.807
|
14c The WI is a Work Task: parent Building Block
|
Trace Management (BB: OAM7-Trace)
WI
Unique_ID
OAM7-Trace
35039
|
566267d8c54c8da6ca65b0a2ce396207
|
32.807
|
32.307 Notification IRP SOAP SS
| |
566267d8c54c8da6ca65b0a2ce396207
|
32.807
|
32.317 Generic IRP Management SOAP SS
| |
566267d8c54c8da6ca65b0a2ce396207
|
32.807
|
32.607 Basic CM IRP SOAP SS
| |
566267d8c54c8da6ca65b0a2ce396207
|
32.807
|
32.617 Bulk CM IRP SOAP SS
| |
566267d8c54c8da6ca65b0a2ce396207
|
32.807
|
32.667 Kernel CM IRP SOAP SS
|
5 Service Aspects
None
6 MMI-Aspects
None
7 Charging Aspects
None
8 Security Aspects
None
|
566267d8c54c8da6ca65b0a2ce396207
|
32.807
|
1 3GPP Work Area
|
Radio Access
X
Core Network
Services
2 Linked work items
OAM&P (Operations, Administration, Maintenance & Provisioning) (Feature: OAM7)
WI
Unique_ID
OAM7
35041
Network Infrastructure Management (BB: OAM7-NIM)
WI
Unique_ID
OAM7-NIM
35042
3 Justification
Circuit is a logic link between two exchange network nodes which bear the user data such as voice, e.g. 64K slot of one 2M E1. Traffic route represents the route via which bearer flow to a specific destination.
To learn the detailed circuit connection relationship between network nodes and traffic route configuration status of the CN CS, bearer transport network related NRM need to be defined, such as circuit, traffic route, etc.
4 Objective
To define Bearer Transport Network (BTN) related NRM which are applicable to CN CS of UMTS.
To specify Bearer Transport Network (BTN) relative NRM definition of the CN CS:
• Specify BTN relative NRM management requirements
• Specify BTN Network Resource Models (NRMs)
• Specify CORBA Solution Set (SS)
• Specify CMIP Solution Set (SS)
5 Service Aspects
None
6 MMI-Aspects
None
7 Charging Aspects
None
8 Security Aspects
None
|
566267d8c54c8da6ca65b0a2ce396207
|
32.807
|
3.2 Principle of radio channel power monitoring
|
The transport channels and physical channels have mapping relations. The following is mapping of transport channels onto physical channels.
25.211
From the figure above, the mapping from transport channel to physical channel is mostly one to one. So, in most cases, we just need to monitor the physical channels.
It is proposed to monitor the power of the following channels:
• Transport channel: FACH、PCH
• Uplink physical channel: DPCH (DPDCH/DPCCH)、PRACH
• Downlink physical channel: DPCH (DPDCH/DPCCH)、CPICH、P-CCPCH、S-CCPCH、SCH、AICH、PICH
In the above channel list, DPCH、PRACH、PDSCH are involved in power control. So, we should record the maximum and mean value of power level for those channels as performance measurements. For other channels, only the configured value will be retrieved.
The following channel power parameters are already present in TS 32.642:
• primaryCpichPower(1)
• maximumTransmissionPower(2)
• bchPower(3)
• primaryCcpchPower(4)
• dlpchPower(5)
• schPower(6)
The following parameters should be added:
• fachPower (7)
• dpchPower(8)
• prachPower (9)
• sccpchPower(10)
• pdschPower(11)
• pichPower(12)
• aichPower(13)
(8)、(9)、(11) are channels which are involved in power control.
4 Objective
Add (7) (10) (12) (13) to TS 32.642
Add (8) (9) (11) to TS 32.403
Update UTRAN Network Resource Model (NRM) Requirements (if needed)
Update UTRAN Network Resource Model (NRM)
Update UTRAN NRM CORBA Solution Set (SS)
Update UTRAN NRM CMIP Solution Set (SS)
Update UTRAN NRM XML format definition
Add UTRAN channel Measurements
5 Service Aspects
None
6 MMI-Aspects
None
7 Charging Aspects
None
8 Security Aspects
None
|
566267d8c54c8da6ca65b0a2ce396207
|
32.807
|
14b The WI is a Building Block: parent Feature
|
(one Work Item identified as a feature)
OAM&P (Operations, Administration, Maintenance & Provisioning) (Feature: OAM7)
Acronym
Unique_ID
OAM7
35041
Technical Specification Group Services and System Aspects TSGS#29(05)0627
Meeting #29, Tallinn, ESTONIA, 26-28 September 2005
Source: SA5 (Telecom Management)
Title: WID WT End-to-end Service Level tracing for IMS (OAM7-Trace-IMS)
Document for: Approval
Agenda Item: 11.27
3GPP TSG-SA5 (Telecom Management) S5-058848
Meeting #43, Bordeaux, FRANCE, 29 Aug - 2 Sep 2005
Work Item Description
Title:
End-to-end Service Level tracing for IMS Unique_ID: 35062
Acronym: OAM7-Trace-IMS
1 3GPP Work Area
X
Radio Access
X
Core Network
X
Services
2 Linked work items
a) Trace Management (SA5 BB: OAM7-Trace)
b) Trace Management, SIP Enhancements for Trace – (CT1 WT : OAM7-Trace-SIP)
WI
Unique_ID
a)
OAM7-Trace
35039
b)
OAM7-Trace-SIP
11046
3 Justification
The Open Mobile Alliance™ (URL:http://www.openmobilealliance.org/) have developed a set of technology agnostic service level tracing requirements, which are now approved [see OMA-RD-OSPE-V1_0-20050614-C.pdf].
The intentions of Service Level Tracing are to improve and simplify end-to-end service diagnostics and to enhance the Mobile Operator’s ability to manage their complex services. Service Level Tracing is aimed at end-to-end service-level diagnostics, rather than per node tracing. By definition, Service Level Tracing is the ability to capture and log all relevant information at each component within a service chain, associated with a specific service that is initiated either by an end user or a component [see OMA-RD-OSPE-V1_0-20050614-C.pdf].
Considering the importance of IMS, 3GPP SA5 SWGD, in coordination with 3GPP CT1, CT3, CT4, will start developing the appropriate specifications for end-to-end service tracing for IMS, and wherever possible to reuse existing 3GPP speciation and their capabilities to fulfil the OMA OSPE service level tracing requirements.
OMA OSPE service level tracing requirements specific to OMA enablers utilising IMS [see OMA-IMSinOMA-V1_0-20050204-C.zip] will be addressed within OMA.
4 Objective
The objectives of this work item are:
a) For 3GPP TSG SA5 to review the OMA requirements for Service Level Tracing and develop their Stage 1, Stage 2 and Stage 3 specifications for Trace as appropriate.
b) For 3GPP TSG SA5, as primary group responsible for Trace in 3GPP, to co-ordinate with 3GPP TSG CT 1, CT3, and CT4 in order for those working groups to develop the specifications under their control that are impacted.
5 Service Aspects
Refer to “OMA Service Provider Requirements” OMA-RD-OSPE-V1_0-20050614-C, The Open Mobile Alliance™ (URL:http://www.openmobilealliance.org/)
6 MMI-Aspects
None
7 Charging Aspects
Refer to “OMA Service Provider Requirements” OMA-RD-OSPE-V1_0-20050614-C, The Open Mobile Alliance™ (URL:http://www.openmobilealliance.org/)
8 Security Aspects
Refer to “OMA Service Provider Requirements” OMA-RD-OSPE-V1_0-20050614-C, The Open Mobile Alliance™ (URL:http://www.openmobilealliance.org/)
|
566267d8c54c8da6ca65b0a2ce396207
|
32.807
|
3 Open Work item status and approval time frame
|
This list reflects the open work items running under the responsibility of TSG SA WG5.
Work items in this colour are closed or building blocks.
|
566267d8c54c8da6ca65b0a2ce396207
|
32.807
|
4 Completed or Terminated Work items
|
This list reflects work items that have been completed or terminated.
Annex A:
Change history
Change history
Date
TSG #
TSG Doc.
CR
Rev
Subject/Comment
Old
New
Nov 2005
S5_44
S5-050529
--
--
Initial draft agreed by SA5#44
Dec 2005
SA_30
SP-050734
--
--
Submitted to SA#30 for Information
Mar 2006
SA_31
SP-060073
--
--
Converted to TR 32.207. Submitted to SA#31 for Information
0.0.3
|
f4da59af4d170d3da6097a508a147acf
|
50.099
|
45.005 New frequency ranges
| |
f4da59af4d170d3da6097a508a147acf
|
50.099
|
45.050 Scenarios for new frequencies
| |
f4da59af4d170d3da6097a508a147acf
|
50.099
|
24.008 Classmark information elements
| |
f4da59af4d170d3da6097a508a147acf
|
50.099
|
45.008 Add frequency ranges
| |
f4da59af4d170d3da6097a508a147acf
|
50.099
|
45.001 Add frequency and channels
| |
f4da59af4d170d3da6097a508a147acf
|
50.099
|
43.030 Add frequency ranges
| |
f4da59af4d170d3da6097a508a147acf
|
50.099
|
43.022 Add channels to be searched
|
2080%
Dec 2002
StartedOngoing
Addition of frequency bands to GSM – Changes for conformance tests
GP-022074
51.010-1 Add testing
0%
Dec 2002
Started
Enhanced Power Control
GP-012748
Realization of Enhanced power control and signaling support
GP-012749
Concept
Changes to 43.051
Changes to 44.004
Changes to 44.018
Changes to 48.058
Changes to 45.001
Changes to 45.002
Changes to 45.003
Changes to 45.008
Nov 2001
Ready for Rel 5. Closed
GERAN MS Conformance test for Enhanced Power Control
GP-012750
• MS test
0%
Dec 2002
Not started
GERAN BTS Conformance test for Enhanced Power Control
GP-012751
• BTS test
0%
Dec 2002
Not started
8PSK AMR HR
GP-012752
Definition of channel coding, performance requirements and signaling support
GP-012753
• Concept
• Changes to 44.018
• Changes to 45.001
• Changes to 45.002
• Changes to 45.003
• Changes to 45.005
• Changes to 24.008
• Changes to 48.058
Jun 2002
Ready for R5. Closed
GERAN MS Conformance test for 8PSK HR
GP-012754
• MS test
0%
Dec 2002
GERAN BTS Conformance test for
8PSK HR
GP-012755
• BTS test
1000%
Dec 2002
GERAN enhancements for streaming services 1
GP-010430
GERAN enhancements for streaming services 1
GP-010430
• Concept
• RLC protocol enhancement (SDU Discard)
Oct 2001
Nov 2001????
Ready for R5. Closed
GERAN enhancements for streaming services 2
GP-010429
GERAN enhancements for streaming services 2
GP-010429
Usage of ECSD
Stage 2
Stage 3
• RLC PDU formats
• MAC header
Jun 2001
Jun 2002
Ready for R5. Closed
Intra Domain Connection of RAN Nodes to Multiple CN Nodes: Overall System Architecture
SA2 Feature
GERAN work for Intra Domain Connection of RAN Nodes to Multiple CN Nodes
GP-020492
Stage 2 (changes to )
• 43.051 Introduction of support for IDNNS in GERAN Iu mode
Stage 3 (changes to )
• 48.016 Use of Gb interface concepts when a network applies IDNNS
• 48.018 Include MSC/VLR identity in CS IMSI paging
Jun 2002
Ready for R5. Closed, accept changes for Gb over IP
700 MHz spectrum support
GP-000449
GERAN support for the 700 MHz band
• Signaling support
• Physical layer definitions
• Receiver performance and RF budget
Ready for R4. Closed
GERAN MS Conformance test for 700 MHz band
GP-000451
• MS test
Jun 2001
Closed
GERAN BTS Conformance test for GERAN interface evolution
GP-000452
• BTS test
0100%
Dec 2002
Ongoing
Real Time QoS for packet services including VoIP
(UTRAN)
HOs: maintenance of real-time QoS while moving between cells in the PLMN including inter-SGSN change and SRNS relocation or possibly other mechanisms (UTRAN)
GP-010431
Handover for the packet switched domain
• Stabile RT handover report 25.936 including header removal
• Update of stage 2
• Update of relevant stage 3 specs
Nov 2001
Closed
Wideband telephony services
(UMTS)
Support of WB AMR in GERAN
GP-000453
GMSK and 8PSK WB FR / HR support
• Channel coding in 45.003
• Signalling for A interface
• Signalling for Iu
• Link adaptation in 45.009
• Receiver performance in 45.005
Apr 2002
Nov 2001
Jun 2002
Ready for R5. Closed
GERAN MS Conformance test for WB AMR
GP-000454
• MS test
0%
Dec 2002
Not started
GERAN BTS Conformance test for WB AMR
GP-000455
• BTS test
1000%
Dec 2002
Not startedOngoing
Location service
(UMTS)
LCS interoprability aspects to GERAN
GP-000456
• Co-ordinated development of GSM LCS Phase 2 and UMTS LCS, S2 and GERAN
Ready for R5. Closed
Location service for GERAN R4
GP-010932
• Work for aligning LCS R4 CN and GERAN
Ready for R4. Closed
Location Services (LCS) for GERAN in A/Gb Mode
GP-011925
• GERAN LCS Stage Two
• Gb interface support for LCS
• L3 protocol support for LCS
• Stage 3 specifications
Feb. 2002
Ready for Rel-5. Closed
Location Services (LCS) for GERAN in Iu Mode
GP-011926
• GERAN LCS stage 2
• Iu interface support for LCS
• Iur-g interface support for LCS
• RRC protocol support for LCS
• Additional impacts on Broadcast of LCS data on packet channels
• Stage 3 specifications
Stage 2-GERAN #8 Feb. 2002
Stage 3 – GERAN #9 Jun 2002
Ready for R5. Closed
GERAN MS Conformance test for LCS
GP-000458
• Develop LCS MS test case work plan (Release 98/99/4)
• Develop LCS MS test cases
?%
Dec 2002 (#11)
Ongoing
GERAN BTS Conformance test for LCS
GP-000459
• Develop LCS BTS test case work plan (Release 99/99/4)
• Develop LCS BTS test cases
?%
Dec 2002
Work has not
started
Uplink TDOA feasibility study
GP-012794
Uplink TDOA feasibility study
GP-012794
• Performing of a feasitibility study
Jun 2002
Closed.
MS Conformance Testing of Dual Transfer Mode
GP-023226
MS Conformance Testing of Dual Transfer Mode
• MS Conformance Testing of Dual Transfer Mode
100%
Feb 2003
Ongoing
Single Antenna Receiver Interference Cancellation (SAIC)
GP-023404GP-023316
Single Antenna Receiver Interference Cancellation (SAIC)
• Determine feasibility of SAIC for GMSK and 8PSK scenarios under realistic synchronized and non-synchronized network conditions. Using a single Feasibility Study, both GMSK and 8PSK scenarios will be evaluated individually.
• Realistic DIR (Dominant-to-rest of Interference Ratio) levels and distributions based on network simulations and measurements.
• Robustness against different training sequences.
• Determine method to detect/indicate SAIC capability.
10%
April 2003
Ongoing
Uplink TDOA location determination for GSM/GPRS
GP-023316
Uplink TDOA location determination for GSM/GPRS
• Addition of U-TDOA in the CS domain
• Addition of U-TDOA in the PS domain
April 2003
Nov 2003
Ongoing
DTM MS test work item to be included in the next revision
Closed work items
New Specifications
GSM No.
TDOC
CR
Subject
CR Comp.
Resp.
TSG
Completion Date
43.051
GERAN overall description
S. Gillaume (Nokia)
GERAN
Nov 00
43.059
Functional Stage 2 Description of Location Services in GERAN
M. Livingston (Nokia)
GERAN
April 2001
44.118
GERAN Iu mode RRC
S. Hamiti
GERAN
June 2002
44.160
GERAN Iu mode RLC/MAC
?
GERAN
June 2002
50.099
TR
GERAN project schedule and open issues
F. Mueller
GERAN
Dec 2002
xx.xxx
TR
Optimized speech in the IMS domain
B. Guarino
GERAN
?
Approved Set on hold #29 Send to SMG #29 CR0000A000 CR has been cancelled
A-1 GERAN radio requirements
A-1.1 Introduction
The GERAN provides a range of bearer services to mobile and stationary users in a variety of application areas and operating environments. The radio access network will be connected to the third generation core network and will as far as possible extend the services of the fixed networks to mobile users.
This document outlines the overall requirements for GERAN release 2000, which includes all GSM/EDGE work items of release 2000. More specific radio requirements, such as radio requirements for the AMR wide band speech codec, are included as references, if available, and are not discussed in this document. The requirements should be used as guidelines for the design of the radio access network.The requirements should be aligned with the requirements on UTRAN.
A-1.2 Definitions and Abbreviations
A-1.2.1 Definitions
GSM/EDGE RAN GERAN is a term used to describe a GSM and EDGE based 200 kHz radio access network. The GERAN is based on GSM/EDGE release 99, and covers all new features for GSM Release 2000 and subsequent releases, with full backward compatibility to previous releases.
A-1.2.2 Abbreviations
3G Third Generation
BER Bit Error Rate
CN Core network
CS Circuit Switched
GERAN GSM/EDGE Radio Access Network
RAN Radio Access Network
RAB Radio Access Bearer
RB Radio Bearer
QoS Quality of Service
PS Packet Switched
UMTS Universal Mobile Telecommunications System
UTRAN UMTS Terrestrial Radio Access Network
A-1.3 High Level Requirements
The following high level requirements have been initially identified for the GERAN in responsibility of SMG2:
• All bearer classes (conversational, streaming, interactive and background) as defined for UTRAN shall be provided
• The same quality of service handling and radio access bearer service attributes shall be supported as required for UTRAN (as described in TS 23.107). Whether the same range of values of the service attributes as supported by UTRAN shall be supported by GERAN in Release 2000 is for further study
• Support for multiple QoS profiles in parallel shall be provided in the GERAN.
A-1.4 Bearer Definition
A-1.4.1 Radio Access Bearers
GERAN shall provide the same radio access bearers as UTRAN. However, voice is foreseen to be important future service and therefor it seen as important to optimize the conversational radio access bearer class for IP voice services.
It is required to have the GERAN support Adaptive Multi-Rate (AMR) CODEC speech and to be consistent with S2 requirements. Further, it is desired to have the GERAN support Tandem Free Operation (TFO) services. Further, voice radio access bearers should be provided with quality and delay comparable to current digital cellular systems.
Figure 1 shows the UMTS QoS architecture. As illustrated in the figure the Radio Access Bearer Service is realized by a Radio Bearer Service and an Iu-Bearer Service.
Figure 1. UMTS QoS architecture.
A-1.4.1.1 Radio Access Bearer Attributes
A set of attributes and their possible values are used to describe a radio access bearer capability. This set has been chosen so that a radio access bearer capability can be entirely defined by giving a value to each attribute of the set. In particular, the set and the associated allowed values enable characterization of future (not yet used or foreseen) transfer needs. For the GERAN the same set of attributes are chosen as for the UTRAN, which are defined in 23.107 [1]. The support of the different values may vary from the radio environment the user is in (indoor, urban, rural and etc.), see section A-1.4.2.1.
The values used by the 3G CN are as follows:
Table 1. Value ranges of the radio access bearer service attributes in UMTS.
Traffic class
Conversational class
Streaming class
Interactive class
Background class
Maximum bitrate [kbps]
<2000 (1) (2)
<2000 (1) (2)
< 2000 – overhead (2) (3)
<2000 - overhead (2) (3)
Delivery order
Yes/No
Yes/No
Yes/No
Yes/No
Maximum SDU size [octets]
<1500 (4)
<1500 (4)
<1500 (4)
<1500 (4)
SDU format information
(5)
(5)
Delivery of erroneous SDUs
Yes/No/-
Yes/No/-
Yes/No/-
Yes/No/-
Residual BER
5*10-2, 10-2, 10-3, 10-4 (6)
5*10-2, 10-2, 10-3, 10-4, 10-5, 10-6 (6)
4*10-3, 10-5, 6*10-8 (6) (7)
4*10-3, 10-5, 6*10-8 (6) (7)
SDU error ratio
10-2, 10-3, 10-4, 10-5 (6)
10-2, 10-3, 10-4, 10-5 (6)
10-3, 10-4, 10-6 (6)
10-3, 10-4, 10-6 (6)
Transfer delay [ms]
80 – maximum value(6)
500 – maximum value (6)
Guaranteed bit rate [kbps]
<2000 (1) (2)
<2000 (1) (2)
Traffic handling priority
1,2,3 (8)
Allocation/Retention priority
1,2,3 (8)
1,2,3 (8)
1,2,3 (8)
1,2,3 (8)
Source statistic descriptor
Speech/unknown
Speech/unknown
Speech/unknown
Speech/unknown
1) Bitrate of 2000 kbps requires that UTRAN operates in transparent RLC protocol mode, in this case the overhead from layer 2 protocols is negligible.
2) The granularity of the bit rate parameters must be studied. Although the UMTS network has capability to support a large number of different bitrate values, the number of possible values must be limited not to unnecessarily increase the complexity of for example terminals, charging and interworking functions. Exact list of supported values shall be defined together with S1, N1, N3 and R2.
3) Impact from layer 2 protocols on maximum bitrate in non-transparent RLC protocol mode shall be estimated.
4) Maximum SDU size shall at least allow UMTS network to support external PDUs having as high MTU as Internet/Ethernet (1500 octets). The need for higher values must be investigated by N1, N3, S1, R2, R3.
5) Definition of possible values of exact SDU sizes for which UTRAN can support transparent RLC protocol mode, is the task of RAN WG3.
6) Values are indicative. Exact values on Residual BER, SDU error ratio and transfer delay shall defined together with S1, N1, N3 and R2.
7) Values are derived from CRC lengths of 8, 16 and 24 bits on layer 1.
8) Number of priority levels shall be further analysed by S1, N1 and N3.
A-1.4.2 Radio Bearers
Mapping of radio access bearers onto radio bearers is up to the RAN as long as the requested QoS is achieved.
Each radio bearer will be mapped to one or more radio interface logical channels for the purposes of transmission over the GERAN. Suggested properties of the GERAN:
• The design of GERAN should allow for several radio bearers to be used simultaneously with single user equipment. This could be used for instance to provide support for multiple QoS profiles in parallel
• The design of GERAN should allow for optimised voice radio bearers in both the PS and the CS domain. The handling of TFO is for further study.
The design of GERAN should allow efficient support of the wide variety of services, including future services, which have yet to be defined.
A-1.4.2.1 Minimum radio bearer capabilities
Giving one of the possible values to each RAB service attribute defines a possible radio access bearer service. However, not all combinations are necessarily supported by the GERAN system. The following table shows potential combinations for the attributes that are expected to change dependent on the radio environment. The values given under the different QoS classes are Maximum bitrate/BER/Max Transfer Delay 1.
Table 2. Minimum radio bearer capabilities.
Operating environment
Propagation conditions
Conversational
Streaming
Background
Interactive
Rural outdoor
(Terminal relative speed to ground up to 250 km/h)
HT100
850/900 Mhz: RA250
1800/1900 Mhz: RA130
T.B.D.
T.B.D.
T.B.D.
T.B.D.
Urban/ Suburban outdoor
(Terminal relative speed to ground up to 120 km/h)
HT100
TU50
T.B.D.
T.B.D.
T.B.D.
T.B.D.
Indoor/ Low range outdoor
(Terminal relative speed to ground up to 10 km/h)
Indoor
TU3
T.B.D.
T.B.D.
T.B.D.
T.B.D.
A-1.4.2.2 RTP/UDP/IP Header adaptation
GERAN shall support header adaptation in order to provide an increase in spectral efficiency. In particular the header adaptation mechanism should not degrade the hand over performance and user perceived quality (e.g. header adaptation mechanism should not degrade the speech quality). Error propagation due to header adaptation should be kept to a minimum or avoided, if at all possible. In addition the header adaptation mechanism should operate under all expected BER and delay conditions.
A-1.5 Handover requirements
This section deals with both intra and inter GERAN handover and cell re-selection requirements. Cell re-selection refers to cell change when in idle mode or ready state, whereas handover refers to change of physical channel (in the same or possibly in a new cell) when in non-idle state.
The overall requirements on GERAN handover and cell re-selection are:
• For support of pre release 2000 terminals the GERAN should provide cell re-selection in the same way as (E)GPRS;
• For support of pre release 2000 terminals the GERAN should provide handover in the same way as GSM;
• Cell re-selection and handover should be in the responsibility of the radio access network2;
• GERAN should support intra- (within a cell) and inter- (between cells) cell handovers;
• For the GERAN release 2000, handover performance should be no worse than for GSM circuit switched services. In particular, the transmission gap should be no more than 150 ms;
• In GERAN release 2000, other requirements related to the HO function shall be of same quality as in GSM release 99 (e.g. neighbourcell measurement rate).
Table on Intra GERAN handover and cell reselection
GERAN R00
PS
GERAN R99
PS
GERAN R00
CS
GERAN R99
CS
GERAN R00
PS
HO
CRS
CRS
No
No
GERAN R99
PS
CRS
CRS
No
No
GERAN R00
CS
No
No
HO
HO
GERAN R99
CS
No
No
HO
HO
HO is for RT services
CRS is for NRT services
„No“ means neither HO or CRS is supported
A-1.5.1 Interworking with other systems
Specific requirements are expected from SA2. The following table should be seen as the working assumption on required handover scenarios between different systems while waiting input from SA2.
Table on Inter GERAN handover and cell reselection
ANSI 136
UTRAN R99 PS
UTRAN R99 CS
UTRAN R00 PS
UTRAN R00 CS
GERAN R00
PS
No
CRS
No
HO
CRS
No
GERAN R00
CS
FFS
No
HO
No
HO
HO is for RT services
CRS is for NRT services
„No“ means neither HO or CRS is supported
A-1.6 Security issues
Specific requirements are expected from SMG10.
A-1.7 Operational requirements
A-1.7.1 Architecture requirements
Specific requirements are expected from SA2.
A-1.7.2 Radio operation environments
GERAN should support all Radio Access Bearers in the radio environments specified in current GSM 05.05.
A-1.7.3 Radio access network planning
For a comparable services, GERAN should provide cell range at least as good as GSM Release 99. GERAN systems should not affect the performance of existing EGPRS/GSM systems.
GERAN should support frequency planning similar to GSM Release 99.
Note: Coverage for RT services of GERAN needs to be defined.
A-1.7.4 Interference Management
GERAN should support interference management at least similar to GSM Release 99. The GERAN solution should not preclude the use of smart antennas.
A-1.7.5 Frequency bands and licensing
GERAN systems should be deployable in at least those frequency bands defined in GSM 05.05 release 99.
A-1.8 Efficient spectrum usage
A-1.8.1 Spectral efficiency
For comparable services, GERAN systems should have significantly higher spectral efficiency as compared to Release 99. It is understood that implementation of increased spectral efficiency may be restricted by the requirement of creating a Release 2000 Standard.
A-1.8.2 Spectrum utilization
For initial deployment GERAN shall support all services in at least 2.4 MHz of spectrum. GERAN shall support all packet domain services (real and non real time) in COMPACT mode deployment. It is recognized that spectrum efficiency may be greater with larger spectrum deployments.
A-1.9 Deployment requirements
A-1.9.1 Deployment
GERAN should be flexible to support a variety of initial deployments.
It should be possible to deploy GERAN with a minimum of upgrades to GSM Release 99 radio equipment. GSM/EDGE RAN may be deployed as a contiguous coverage, Island coverage, or Spot coverage system. It is anticipated that GERAN will also be deployed on a city-by-city basis.
A-1.9.2 Backward compatibility
It should be possible to deploy GERAN in spectrum shared with GSM Release 99, as well as other GSM systems. GERAN should be deployable in carriers and time slots adjacent to those supporting GSM Release 99, at least with fixed division of time slots between GERAN and the other systems.
It is recognized that there may be advantages to dedicating radio resources system-wide to some types of GERAN operation.
A-1.9.3 Complexity / cost
It should be possible to provide a variety of MS as well as Base Station types of varying complexity, cost and capabilities in order to satisfy the needs of different types of operator and user scenarios. The Release 2000 is expected to imply the same RF properties as a Release 1999.
A-1.9.4 Terminal
GERAN systems should support a variety of terminal types, including advanced feature phones, PDA’s, PCMCIA cards, and other terminal types. Hand portables and PCMCIA card sized GERAN terminals should be optimized in terms of size, weight, operating time, range, and the effective radiated power and cost/performance ratio.
A-1.9.5 Network
For further study
A-1.10 Requirements from bodies outside SMG
A-1.10.1 Electromagnetic compatibility
GERAN systems should cause no more interference to other equipment than current GSM-based systems.
A-1.10.2 RF radiation
GERAN systems should operate at RF emission power levels consistent with applicable recommendations and specifications for electromagnetic radiation.
A-1.10.3 Security
For further study
A-1.11 Evolution of GERAN
Release 2000 of GERAN should include efficient support of RT services in the PS domain and it should be aligned with UMTS. The GERAN shall be defined so that it can be implemented in phases with increasing functionality (for example making use of new technology), while allowing the maximum possible backwards compatibility. The introduction of new functions should be done in a manner that maximizes forward compatibility with enhancements expected in subsequent releases. The definition of GERAN should allow evolution to higher bit rates.
A-1.12 Open Issues
This section summerizes the open issues that have been identified in this document.
1. Is there support for multiple QoS profiles in parallel in R99
2. A discussion on the relation of TFO to the Transcoder (TRAU) position in the architecture highlighted the issue of how UTRAN deals with TFO. The following questions arose:
1. Clarification on how TFO is handled in UMTS (This is a question for 3GPP TSG S4))
2. What voice requirements will come from S2
3. Input from SA2 is expected on the RAB attribute value ranges.
4. The T.B.D. in table 2 need to be resolved. Another open issue in the table is whether other propagation models should be included, e.g. BUx.
5. Verify that the speech gap during handover should be no more than 150 ms is a GSM requirement.
6. The delay and data loss requirements on different handovers and cell re-selection shall be specified further. The requirements depend on the service and that should be reflected as well.
A-1.13 References
[1] TSG SA2, 23.107, “QoS Concept and Architecture”.
A-2 History
Document history
23th February 2000
First draft (V0.0.1)
2nd April 2000
Updated after GERAN #1 and EDGE WS #13 (V0.0.2)
8th May 2000
Updated after SMG2 #35 (V0.0.3)
22nd May 2000
Updated after SMG2 GERAN WS #2 (V0.0.4)
24th May 2000
Updated during SMG2 #36 (V0.0.5)
2nd August 2000
Updated for 3GPP S3 meeting (V0.0.6)
28th August 2000
Updated after SMG2 GERAN release 2000 and beyond Adhoc #1
9th October 2000
Updated after TSG GERAN #1 as 50.099 (V0.0.1)
6th November 2000
Updated after TSG GERAN Adhoc on release 2000 and beyond #2 as 50.099 (V0.0.2)
12th February 2001
Updated after TSG GERAN #3 (V0.0.5)
April 2001
Updated for TSG GERAN #4 (V0.06)
7th May 2001
Updated for TSG GERAN Adhoc on released 2000 and beyond #5 (V0.07)
11th May 2001
Updated during TSG GERAN Adhoc on release 2000 and beyond #5 (V0.08)
28th May 2001
Updated for TSG GERAN #5 (V0.09)
27th August 2001
Updated for TSG GERAN #6 (V0.10)
26th Nov 2001
Updated for TSG GERAN #7 (V0.11)
30th Nov 2001
Updated for TSG GERAN #7 (V0.12)
30th Nov 2001
Updated for TSG GERAN #7 (V0.13)
15th April, 2002
Updated for TSG-GERAN #9 (v0.15)
19th April, 2002
Updated for TSG-GERAN #9 (v0.16)
24th June 2002
Updated for TSG-GERAN #10 (V0.17)
28th June 2002
Updated during TSG GERAN #10 (V0.18)
25th August 2002
Updated for TSG GERAN #11 (V0.19)
30th August 2002
Updated during TSG GERAN #11 (V0.20)
20th November 2002
Updated during TSG GERAN #12 (V0.21)
Editor: Frank MuellerDavid Bladsjö, Ericsson
Email: [email protected]
Tel: +46-8-7570287 4048657
|
e43c8396f49910c64271dab0534b85fe
|
23.876
|
1 Scope
|
This technical report defines methods for supporting push services by 3GPP bearers. The mechanisms defined apply to existing bearers for the 3GPP Packet Switched Domain (PS domain), Circuit-Switched Domain (CS Domain), and the IP Multimedia Core Network Subsystem (IMS), Multimedia Broadcast Multicast Service, and Wireless LAN. This technical report addresses the requirements for supported push services as defined in 3GPP TS 22.174 Push Service; Service aspects (Stage 1). Any necessary changes identified during this work will be introduced by means of CRs to the appropriate specifications.
Definition of push functions that apply to push application servers is outside the scope of this work. The definition of push functions that are best implemented in push application servers such as a Push Proxy and Push Initiator will be undertaken by other standards bodies and industry forums.
|
e43c8396f49910c64271dab0534b85fe
|
23.876
|
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 22.060: " General Packet Radio Service (GPRS); Service description; Stage ".
[3] 3GPP TS 22.174: " Push Service; Service aspects (Stage 1) ".
[4] 3GPP TS 23.039: " Interface protocols for the connection of Short Message Service Centres (SMSCs) to Short Message Entities (SMEs) ".
[5] 3GPP TS 23.040: " Technical realization of the Short Message Service (SMS) ".
[6] 3GPP TS 23.060: " General Packet Radio Service (GPRS); Service description; Stage 2 ".
[7] 3GPP TS 23.228: " IP Multimedia (IM) Subsystem - Stage 2 ".
[8] 3GPP TS 23.002: "Network Architecture".
[9] 3GPP TS 29.007: "General Requirements on interworking between the PLMN and the ISDN or PSTN".
[9] 3GPP TR 23.875974: "Support of Push Services".
[10] 3GPP TR 23.910: "Circuit Switched Data Bearer Services".
|
e43c8396f49910c64271dab0534b85fe
|
23.876
|
3 Definitions, symbols and abbreviations
| |
e43c8396f49910c64271dab0534b85fe
|
23.876
|
3.1 Definitions
|
For the purposes of the present document, the terms and definitions given in 3GPP TS 22.174 [3] and the following apply.
delivery network: a network that supports connectionless or connection oriented push services. A delivery network may simply be a GPRS network.
application server: a server that provides push services through a delivery network, e.g. via an IP connection
user IP address: an IP address provided by the delivery network that can be used by an application server to for access to a push services user. The address may be permanently assigned (static) or temporarily assigned (dynamic).
user-ID: an identity or name that can be used to deliver push content to a user in a delivery network The format of user-ID is dependent on the protocol for the push services.
user availability: the ability of an delivery network to provide push service to a subscribed user.
user terminal: the end user equipment that receives push content.
long-lived PDP Context: this is a PDP Context that remains active/open for an indefinite period of time. Also referred to as “always-on PDP context”.
always-on PDP Context: this is a PDP Context that remains active/open for an indefinite period of time. Also referred to as “long-lived PDP context”.
Push Data: data sent by the push initiator to the push recipient, of a format known to the receiver (push recipient), and not otherwise defined by the push service.
Push function: the function in the PLMN that receives the Push Data from the Push initiator. The push function is responsible for delivering the push data to the Push recipient. The Push Function may also be referred to as a Push Proxy or Push Proxy Gateway.
Push initiator: the entity that originates push data and submits it to the push function for delivery to a Push recipient. A Push initiator may be e.g. an application providing value added services.
Push recipient: the entity that receives the push data from the Push function and processes or uses it. This may include the UE with which the PLMN communicates with, the user agent with the application level address, and the device, machine or person which uses the push data. A Push recipient is controlled by an individual user .
Push service: a service capability offered by the PLMN. The Push Service is initiated by a Push Initiator in order to transfer push data (e.g. data, multimedia content) from the Push Initiator to the Push Recipient without a previous user action. The Push Service could be used as a basic capability or as component of a value added service.
Push User agent: is any software or device associated with a Push recipient that interprets Push Data to the user. This may include textual browsers, voice browsers, search engines, machine or device interface software, etc.
|
e43c8396f49910c64271dab0534b85fe
|
23.876
|
3.2 Symbols
|
For the purposes of the present document, the following symbols apply:
|
e43c8396f49910c64271dab0534b85fe
|
23.876
|
3.3 Abbreviations
|
For the purposes of the present document, the following abbreviations apply:
NAT Network Address Translator
NRPCA Network Requested PDP Context Activation
OTA Over The Air delivery protocol
PP Push Proxy
PI Push Initiator
|
e43c8396f49910c64271dab0534b85fe
|
23.876
|
4 Architecture Requirements
| |
e43c8396f49910c64271dab0534b85fe
|
23.876
|
5 Push Architecture Overview
|
The Push Service Architecture overview is shown in figure 1. This includes the push application servers Push Function (or Push Proxy) and Push Initiator as well as the bearer services available as the delivery network and the Push Recipient or UE. The definition of functions in the Push Function (Push Proxy) and Push Initiator are outside the scope of this TR. Figure 1 also shows the Push Function performing bearer selection, the definition of how this is performed and the criteria for bearer selection are part of the definition of the Push Function and are outside the scope of this TR. Figure 1 depicts the Push Function being located within the PLMN, this is a logical representation of the Push Service Architecture and does not imply the physical co-location of a Push Function within the PLMN infrastructure.
The definition description of the delivery network (bearers) used to support push services and how those bearers are established, maintained and withdrawn is the main focus of this sectionTR.
Figure 1: Push Service Architecture Overview.
|
e43c8396f49910c64271dab0534b85fe
|
23.876
|
5.1 Push Bearers in the PS Domain
|
This section describes the use of various mechanisms in the PS Domain to establish and/or maintain a bearer service connection to the UE over which Push services may be delivered.
Editors Note: the following bullets provide guidance for further work
• Push using Long Lived PDP Context
◦ This section describes the use of an existing PDP Context for Push services.
▪ Push with Static IP Address Assignment
◦ This section describes the use of 3GPP TS 23.060 section 9.2.2.2 Network-Requested PDP Context Activation procedure to establish and carry Push services to a UE.
• Push with Dynamic IP Address Assignment
◦ This section describes a mechanism to establish a PDP Context that can be used to carry Push services to a UE when the PS Domain implements Dynamic IP address assignment.
• Push using SMS in PS Domain
◦ The section describes how Push services can be delivered to a UE using the services defined for Short Message Service in the PS Domain.
|
e43c8396f49910c64271dab0534b85fe
|
23.876
|
5.2 Push Bearers in the CS Domain
|
This section describes the use of various mechanisms in the CS Domain to establish and/or maintain a bearer service connection to the UE over which Push services may be delivered.
Editors Note: the following bullets provide guidance for further work
• Push over Circuit-Switched Data Bearer
◦ This section describes the use of a circuit-switched data bearer to deliver Push services. The Circuit-Switch Data connection is established based on the mechanisms described in 3GPP TR 23.910 Circuit-Switched Data Bearer Services and 3GPP TS 29.007 General Requirements on interworking between the PLMN and the ISDN or PSTN, section 9.2 Data Calls.
• Push using SMS in CS Domain
◦ The section describes how Push services can be delivered to a UE using the services defined for Short Message Service in the CS Domain.
|
e43c8396f49910c64271dab0534b85fe
|
23.876
|
5.3 Push in the IP Multimedia Subsystem
|
Editors Note: the following bullets provide guidance for further work
• Push using SIP
◦ The solution described in this section defines a method using the SIP protocol in IMS to carry Push services to a UE.
|
e43c8396f49910c64271dab0534b85fe
|
23.876
|
5.4 Push using MBMS
| |
e43c8396f49910c64271dab0534b85fe
|
23.876
|
5.5 Push using WLAN
| |
e43c8396f49910c64271dab0534b85fe
|
23.876
|
6 Analysis and Conclusion
|
Annex <A> (normative):
<Normative annex title>
Annex <B> (informative):
<Informative annex title>
Annex <X> (informative):
Change history
Change history
Date
TSG #
TSG Doc.
CR
Rev
Subject/Comment
Old
New
200301
Created at SA2#29 San Francisco
0.0.0
2003-01
Revised at SA2#29 San Francisco
0.0.1
|
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
1 Scope
|
The present document summarises the results of the feasibility study about inclusion of Wideband Distribution Systems, or WDS, as part of third generation networks in 3GPP standards.
It covers the history of the proposal, shows and discusses the results of relevant analysis and simulation activities, highlights the possible degree of integration to the current status of 3GPP network architecture and concludes with the indication for a possible way forward in the standardisation path.
|
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
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] End to End Simulations Detail, TEK 3GPP ETE SD v.0.1, R4-011578 (RAN WG4 #20).
[2] “Universal Mobile Telecommunications System (UMTS); UE Radio Transmission and Reception (FDD) (3GPP TS 25.101 version 4.1.0 Release 4)”, Ref. RTS/TSGR-0425101Uv4R1, July 2001
[3] “Universal Mobile Telecommunications System (UMTS); UTRA (BS) FDD; Radio transmission and Reception (3GPP TS 25.104 version 4.1.0 Release 4)”, Ref. RTS/TSGR-0425104Uv4R1, July 2001
[4] ADS simulation model parameters TEK 3GPP 21_ADS, R4-011579 (RAN WG4 #20)
[5] Technical justification and overall advantages for UTRA Wideband Distribution Subsystems (WDS) R4-010668 (RAN WG4 #17).
[6] Comments on WDS O&M impact from RAN3 and SA5, R4-011609 (RAN WG4 #20)
|
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
3 Definitions, symbols and abbreviations
| |
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
3.1 Definitions
| |
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
3.2 Symbols
| |
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
3.3 Abbreviations
|
For the purposes of the present document, the following abbreviations apply:
3GPP Third Generation Partnership Project
ACLR Adjacent Channel Leakage Ratio
ACS Adjacent Channel Sensitivity
ALC Automatic Level Control
BS Base Stations
CPICH Common PIlot CHannel
DL Downlink
FDD Frequency Division Duplex
GPS Global Positioning System
ID Identification
IPDL Idle Period for Down Link
Iub Interface between RNC and BS
LCS Location Services
LLI Low Level Interface
LMU Local Measurement Unit
LNA Low Noise Amplifier
MCPA Multi Carrier Power Amplifier
O&M Operation and Maintenance
QoS Quality of Service
OTDOA Observed Time Difference of Arrival
PCF Position Calculation Function
RF Radio Frequency
RNC Radio Network Controller
TDD Time Division Duplex
TR Technical Report
UE User Equipment
UDD Unconstrained Delay Data bearer service
UL Uplink
UMTS Universal Mobile Telecommunication System
UTRAN UMTS Terrestrial Radio Access Network
WCDMA Wideband Code Division Multiple Access
WDS Wideband Distribution System
WI Work Item
4 Definition of WDS in FDD UTRAN
4.1 Introduction
UTRAN FDD Base Stations can sometime include Ancillary Equipment like masthead amplifiers or remote radio heads, that may add flexibility and reduce cost of installation.
These solutions are embedded in BS as ancillary RF amplifiers and are therefore seen as integral part of it in a single-vendor deployment scenario.
In order to improve flexibility of radio access network solution, a new type of equipment is proposed, here called Wideband Distribution System, or WDS.
WDS are altogether similar devices, capable of remotisation of BS RF interface, but offering flexible and multiple RF interface to one or more BS or sub-equipped BS. The so-defined WDS shall include one or multiple RF front-ends, RF transmission, and interfaces capable of supporting one or multiple BS.
WDS may be designed to operate in any full UTRA FDD paired or TDD bands according to regional requirements.
Similar definitions are possible in TDD scenarios and may be considered as part of further work, and analysis of the TS 25.105 UTRA (BS) TDD specification has not highlighted any parameters that suggest WDS would not comply in a TDD environment.
The degree of performance impact shall be assessed in this document to understand the effect on multi-carrier WCDMA signals in order to maintain compliance to the relevant standard in the coverage area. The test and simulation scenarios in this report are made with the assumption that of no impact from any passive distribution system. Therefore the results are of an ideal nature and may need to be adjusted to suit the class of base station utilised for deployment.
|
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
4.2 WDS Architecture
|
The key attribute of WDS is its capability to enable the radio interface of a number of BS to be remote, and hence support a distributed multi-carrier and multi-operator network.
WDS is in general an active device and includes, but is not limited to, one or multiple RF front-end (LNA, MCPA) and RF transmission interfaces capable of supporting one or multiple BS. Other ancillary functions may be included as required for best system integration. It also includes O&M facilities and interfaces in order to fulfil any supervision requirements.
Figure 1 Network Configuration with WDS
WDS generally include a number of functions that are required for correct operation. Those functions are listed and described in Table 1 here below:
Function Definition
Function Description
Centralised multi-operator RF interface
It provides RF independent interface to multiple BS belonging to different networks. It must include provisions such as RF isolation (>30dB), power threshold detectors, and ALC to prevent that any malfunctioning at one network may affect other networks. Transmit RF power at BS interface can be as low as a fraction of a Watt
Transmission
It provides the proper wideband link to a number of remotely placed sites that host suitable RF amplifiers and other devices (RF front-end)
RF transmit MCPA
It amplifies all available RF channels in the downlink direction, and therefore shall offer suitable in a multi-carrier scenario.
Power classes can be defined on a wide range, and amplifier technology shall accordingly change to maintain best efficiency
RF receive LNA
It amplifies uplink signals before they are fed back to BS receivers. Its dimensioning is basic in order to optimise uplink dynamic range (NF, Intermods, Blocking)
RF filtering/diplexing
It provides a common TX/RX antenna connector at the remote site, and includes proper selectivity for achieving interference protection as required in the various deployment scenarios
Other functions
Other functions shall be included if applicable, e.g. diversity and O&M
Table 1 – WDS functions breakdown
|
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
4.3 Practical Deployment Examples of WDS
|
Two examples are given only for information and topology understanding purposes of practical deployment for WDS, with reference to previous functions listing in Table 1. Additional indications will require further study on a dedicated Technical Report on WDS Deployment Scenario.
4.3.1 In-building Deployment
A standard layout is shown, where 2 centralised BS belonging to 2 different networks feed a WDS with 24 Remote Low Power RF Heads, with all RF channels distributed throughout the building (or part of it) in one single cell. The case could be easily expanded to a higher number of networks without affecting the concept.
In this deployment case coverage antennas are often fed by means of a small passive distribution network, i.e. 5 to 20 metres coaxial cable and RF splitters/couplers, for maximum losses in the range of 10dB at UMTS frequencies. This loss adds up to the RF path loss. The number N of Remote Heads being conveyed on a single sector would increase UL NF by 10 Log N, and this will require further consideration in possible WDS specification and deployment scenarios.
4.3.2 Outdoor Deployment for Small Cells
A possible layout is shown, where 2 centralised BS feed a WDS with 1 Remote RF Head, with dedicated RF channels to each of the small outdoor cells that may be possible with WDS. The case could be easily expanded to a higher number of networks without affecting the concept.
In this deployment case receive diversity path has been added. Hence uplink time delay and RF gain inequality control must be considered as a possible requirement.
|
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
4.4 Benefits of WDS
| |
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
4.4.1 Technical and Deployment Aspects
|
Because of its very principle, WDS may bring technical and economical advantages as summarised here:
1. Possible BS and RNC co-location in centralised equipment locations allowing shared facilities, increased implementation flexibility and trunking efficiency.
2. Distributed RF wideband microcellular heads, with lower RF transmission power to cope with most stringent environmental compatibility and scalable traffic capacity requirements.
3. Better and easier flexibility in network planning and upgrading, and on capacity and location systems implementation
4. Sharing opportunities, leading to cost reduction and reduced visual impact for cell sites with the added possibility of increased protection from co-channel and adjacent channel (intra-networks) interference
5. Faster and easier network rollout and maintenance in currently established transmission infrastructures
6. Enabling network manufacturers to shipping base stations and other network elements more quickly
|
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
4.4.2 Standardisation Aspects
|
A standardisation process of WDS in the UMTS UTRA scenario is envisaged for:
1. Removing technical uncertainty risks from Operators and prevent from integration surprises
2. Reducing the burden of additional responsibility for Operators in defining their own specifications for WDS
3. Enabling fulfilment of EU recommendations on network sharing in specific scenarios by providing a common radio distribution solution
|
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
5 RF Feasibility Study
|
5.1 RF Performance Discussion
Starting from reference [5], the RF feasibility study addresses the technical analysis of WDS performance based on its non-linear characteristics that may affect RF transmission through it, and derived as a function of a set of RF parameters at its interface with BS.
The following tables, extracted from the reference, may be useful to highlight why downlink behaviour is more critical than uplink, as this is the primary justification for the simulation work that has been carried out.
BS on TS 25.104
WDS
Repeater on TS 25.106
Transmit path (DL)
116dB
127dB
109dB
Measured as distance in dB between output power and background noise in 1Hz bandwidth for comparable output power levels in the three cases:
Po_Node B = 31dBm
Po_WDS = 32dBm
Po_Rpt = 30dBm
Receive path
(UL)
129dB
143dB
91dB
Measured as distance in dB between input noise level in 1 Hz bandwidth and blocking level – BS, WDS and Repeaters are assumed with NF= 5dB
Table 2 - Dynamic Range Comparison
BS interface
WDS transmission
WDS RF Head
<<1 microSec
5 microSec/Km*
<<1microSec
Note *: length of physical transmission medium, velocity factor = 0.65
Table 3 - WDS Incremental Time Delay (worst case)
5.2 Simulation Program and Assumptions
The study is based on simulations carried out on a suitable simulation program that takes into account non-linear parameters of WDS and is capable of simulating an equivalent stimulus for one or more WCDMA Base Stations.
Non Linear parameters of WDS are defined based on currently and realistically available technologies.
5.3 Definition of RF Power Levels
For the scope of simulating overall performance, WDS RF transmit power levels utilised in the simulation were focussed mainly on the higher output power level (43dBm) specified in TS25.104. Some tests were carried out at the lower power levels of 20dBm and 33dBm in order to give an indication of likely performance with the yet undefined lower power BS classes.
5.4 RF Simulation Parameters
The simulation programme was focussed on the downlink parameters which were highlighted as the primary concern for system feasibility. ACLR was identified as the main area of focus, in the single and multi-carrier scenario. Other parameters of interest included Spectrum Emissions (Out of Band) and Measurement of Occupied Bandwidth.
In order to carry out initial simulations and attain RF parameters, a realistic WDS system model (Test model 1) was created for single carrier scenarios. The model assumed connection to a compliant BS as supplied in the software package. The model utilised for initial ACLR testing with output power of 20dBm and 33dBm is shown in Figure 2 below.
Figure 2 Test model 1 of the WDS
A more complex model (Test Model 2) was then created in order to individually model the non-linearities of each component and reflect values found from actual measurements on real world devices.
The model block is shown below in Figure 3.
Figure 3 Test model 2 of the WDS
Greater detail of the test model parameters are available in reference [4].
|
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
5.5 Simulation Results
| |
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
5.5.1 ACLR Testing
|
The simulations are carried out over a range of ACLR levels in order to assess the degradation attributable to the inclusion of a WDS and to attain the minimum ACLR input values to achieve the current specified output levels according to TS 25.104.
The performance of the BS was adjusted in order to achieve the range of ACLR figures around the specification levels. This enabled the ACLR figures to be varied whilst the output power remained constant.
Simulations were first taken without connecting WDS in order to attain the Input ACLR, and then repeated with WDS connected in order to attain the Output ACLR figures.
5.5.1.1 WDS Output Power 20dBm
The BS model was set with a constant gain of 20dB, whilst attenuation of 25.36dB was inserted in order to attain an input level of –2dBm to the WDS, thus achieving a system output power of 20dBm. Test Model 1 was used for this simulation.
Graphs 4 – 7 show the results found.
Figure 4
Figure 5
Figure 6
Figure 7
5.5.1.2 WDS Output Power 33dBm
The BS model was set with a constant gain of 20dB, whilst attenuation was decreased to 12dB in order to increase the input level to the WDS to 11dBm, thus achieving system output power of 33dBm.
Graphs 8 – 11 show the results found.
Figure 8
Figure 9
Figure 10
Figure 11
5.5.1.3 WDS Output Power 43dBm
Test Model 2 was utilised for this scenario, with different parameters related to a higher power WDS. Input ACLR was varied by means of adjustment of the performance of the BS in the same manner as in the previous set-up.
The results found are shown in figures 12 – 15 below.
Figure 12
Figure 13
Figure 14
Figure 15
|
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
5.5.2 Multicarrier ACLR
|
In order to assess the impact of a multi-operator environment where multiple BS’s were connected to WDS, a test simulation was carried out with two BS, transmitting on adjacent carrier frequencies, connected to WDS Test Model 2 and the ACLR was measured. The test setup is shown in Figure 16 below.
Figure 16 Test setup for multi-operator ACLR.
The gain of the test set-up shown resulted in output power of 23dBm per channel.
The ACLR figures shown below were measured as specified in TS 25.141 for multi-carrier tests. The first test utilises the test bed shown in figure 16 with dotted direct connection. The measurement gives therefore an indication of a multicarrier scenario without the added performance impact which WDS introduces.
Test 2 (as shown in figure 16) was the same in all ways to Test 1 but with the WDS Test Model 2 included in order to record the performance impact. Only two measurement points were simulated due to extremely long simulation runs.
The results are shown in table 4 below.
ACLR offset
+10 MHz
+5MHz
-5MHz
-10MHz
A
B
A
B
A
B
A
B
BS Direct connection
64.24
53.40
54.81
45.72
54.80
45.91
64.42
54.01
With WDS
64.09
50.05
54.54
43.13
54.43
43.80
64.07
50.10
Table 4 Results for multicarrier ACLR
The BS output ACLR in the example [columns A] is that of a very high quality signal and it can be seen that the impact of WDS on this signal is minimal, whilst [columns B] reflect a TS 25.104 compliant BS and the impact of WDS is clearly visible. Simulation results show that the WDS ACLR in the multi-carrier scenario is similar to the single-carrier scenario.
|
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
5.5.3 Measurement of Occupied Bandwidth
|
WDS test model 2 was used to measure the occupied bandwidth resulting from the inclusion of the WDS. The occupied bandwidth containing 99.5% of the integrated transmitted power (43dBm) was measured at 3.855MHz (see figure 17).
Figure 17 Measurement of the occupied bandwidth.
Results from Figure 17.
Lower Side %
Upper Side %
Occupied Bandwidth (MHz)
Total Power (dBm)
0.242
0.241
3.855
43.237
|
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
5.5.4 Spectrum Emission Mask ( Out of Band Emissions )
|
The out of band emissions were measured as specified in TS25.141 for a single carrier scenario with output power of 43dBm using WDS model 2.
The result found is shown in the graph below (figure 18) along with the relevant mask for Power Out greater than or equal to 43dBm (reference TS25.141 6.2.5.1)
Figure 18 Measurement of the spectrum emission and relevant mask.
|
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
5.6 Simulation Summary
|
The simulation programme and hence the results set out to demonstrate two specific areas of interest.
1. The range of input levels available to a system which would feed a WDS system and its relationship to current specifications.
2. The margin required with the introduction of WDS over a number of output power classes.
|
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
5.6.1 ACLR
|
It was found that a 3dB margin is typically required to accommodate the effects of the inclusion of a WDS to an existing Base Station set-up.
There is a minor difference in performance which is dependent on WDS input signal level. With low input power signal levels, as demonstrated in the example with 20dBm output power, the ACLR at +/- 10MHz becomes “non-linear” at high ACLR input levels. This effect is due to the closer proximity of the input signal level to the noise floor.
A similar effect is evident for higher input power levels although this is found on the +/-5MHz area as the signal levels are subjected to WDS amplifier non-linearity.
Two simulation runs were carried out to understand the impact of multi-carrier operation. The results indicated that for two carriers, each with output power of 23dBm, the impact on ACLR is similar to that seen with a single carrier.
|
198feca7a70995985a5cc7f203dbedb2
|
25.867
|
5.6.2 Occupied Bandwidth / Out of Band Emissions
|
Single carrier testing with WDS output power of 43dBm show that there is negligible impact on the occupied bandwidth or output emissions due to the inclusion of WDS.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.