Requirements and Applications of IMS for the SIP protocol

Summary

In recent years, the IP Multimedia core subsystem (IMS) has increasingly become an ideal solution for the integration of fixed networks and mobile networks. Session Initiation Protocol (SIP) is the basic control protocol of IMS, its own characteristics make it play an increasingly important role in the process of moving fixed network and mobile network towards the next generation network (NGN. This article analyzes the basic application and extended application of SIP in IMS, and describes the application of SIP in IMS through the typical process of sip signaling in IMS.

1. Brief Introduction to IMS and SIP protocols

The IP Multimedia core subsystem (IMS) is a subsystem proposed by the third-generation mobile communication partner project (3GPP) to support IP multimedia services. It is characterized by the adoption of Session Initiation Protocol (SIP ), communication has nothing to do with the access mode. It provides multiple media services, including separation of control functions and bearer capabilities, separation of call and session, separation of applications and services, separation of services and networks, and integration of mobile and Internet services. IMS conforms to the development trend of communication network convergence.

SIP is designed based on the two most successful services web and e-mail on the Internet. Drawing on Internet standards and protocol design ideas, adhering to the principles of simplicity, openness, scalability, and reusability, it provides a method to combine simple applications into complex services to form multimedia communication networks and provide multimedia services. SIP establishes and controls various types of point-to-point media sessions in a convenient way. Similar to the Internet protocol, it adopts a modular structure, request/response mode, and text-based mode. Therefore, it is very simple and flexible to use, and is easy to upgrade and expand. SIP consists of basic SIP protocols and a series of SIP extensions for mobile businesses. The basic protocol of SIP is defined by the request instruction document (RFC) 3261 of the Internet Engineering Task Group (IETF), and the SIP Extension is composed of a series of RFC documents, including rfc3455, rfc3311, rfc3262, and RFC
More than 20 documents such as 3325.

 2. SIP protocol in IMS

Due to the flexibility of SIP, 3GPP adopts sip as the session control protocol in R5 to design IMS. 3GPP does not define a new SIP protocol, but simply uses the SIP defined by IETF in some way. Therefore, there are also specific requirements for SIP in public mobile networks, such as low bandwidth, roaming, security requirements, service quality (QOS), and billing control.

In the SIP model, to establish a session, the user proxy client initiates a request to the user proxy service. Requests are routed through the proxy server in the network. In addition, to provide the location information of the user agent, you need to map the SIP address to an IP address. This model is selected for the 3GPP IMS architecture.

The user agent in IMS is the user device (UE ). The proxy server and registration server in IMS are the network entities of the call session control function (cscf. There are three types of cscf:

1) Service cscf (S-CSCF), act as a registration server and activate application business control based on user data;

(2) The proxy cscf (P-CSCF) is the first contact point of UE in IMS network; the sip signaling message is transmitted between P-CSCF and UE;

3) Querying cscf (I-CSCF) is the first contact point for an external network, especially for an external IMS network.

In IMS, the following problems need to be solved:

1) use wireless interfaces effectively

Wireless interfaces are rare resources. Therefore, the exchange of signaling messages between mobile terminals and networks must be minimized. All developed mechanisms must make the use of wireless interfaces the most effective.

2) Minimized terminal support

Because the terminal should be a relatively small device with memory requirements, energy consumption and processing capabilities, it should be minimized.

3) Roaming and non-roaming

All requirements may be roaming or non-roaming. In both cases, the signaling process will not change significantly.

4) terminal Mobility Management

Because terminal mobility is managed by the access network, terminal mobility management is not required in SIP.

5) IPv6

3gppims is designed separately for IP version 6. Therefore, all protocols must support IP version 6.

6) sip outbound Proxy Server

The SIP outbound proxy server is used to support roaming and non-roaming scenarios. The SIP outbound Proxy Server can be located either in the VPC or in the visiting network. There must be a comprehensive mechanism for the mobile device (UA) to learn the address of the SIP exit proxy server.

7) Registration

One or more sip registration servers must be maintained for one private network. The SIP registration server identifies the user and registers the IP address of the user. Once an end user is activated, UA reads its configuration data. This data can be stored in the SIM card or any other storage device. The configuration data contains the identification of a VPC. The device finds the SIP registration address from the domain name of the network. The terminal sends a registration message through the SIP exit proxy server. To support registration search, the private network must contain one or more sip servers. These are the edge proxy servers of the network. Their mission is to serve as the first point to connect to the VPC. With the help of locating the server, they decide which sip registration server to assign to a specific user. Regardless of whether the UA is roaming or not, the Registration Program is the same.

A) registration is required. You must register with IMS before receiving any session invitation. In addition, you must register before starting a session. To send SIP requests to sessions and services that are considered to be standard, the SIP service proxy in the private network needs to know when and which end user is legal. The user can be identified early, so that the authentication will not lead to a delay in the quick dial. The user is assigned to a specified service proxy. The Service proxy downloads the service profile to trigger the service. Therefore, the 3GPP delegate UA to register before starting the session.

B) valid registration. Due to the scarcity of wireless interface resources, each registration must be valid to ensure that UA is accessible to both the home network and the home network.

8) Cancel User Registration

You must have a program to cancel registration from the network. This program can be used. If you disable the terminal, a register with a suspension time of 0 will meet this requirement.

9) User Identification

A) private user identification. To use 3 gppims, a user is assigned a private user identity. The owner network is assigned a private user identity, which is used to determine the uniqueness of the user in a network. For example, a private user identity is used for authentication, authorization, and management (AAA ). The private user identity is not used to route the SIP message. A private identity is expressed as a network access identifier (NAI), which is defined in rfc2486.

B) private user ID used for registration. The UA must send a private user identity to the SIP outbound proxy server and the register. Private user identity is the basic proof during mobile user registration. In order to use 3 gppims, a user is assigned to one or more public user identities. When a user requests to communicate with other users, the user uses the public user identity. A user can have different shapes, each of which contains different public user identities. The format of the public user identity is the unified Resource Identifier (URI) of the SIP ).

10) sip route

A) sip outbound proxy server. The 3GPP architecture includes a sip outbound proxy server, which is generally configured in the visiting network. This outbound proxy server provides local services such as SIP Message compression and security functions. In addition, the exit server can work with the media Reservation Mechanism to provide authentication and authorization support for media appointments. All mobile terminals that initiate a session establishment attempt must pass through the exit proxy server so that the services provided by the exit Proxy Server can be sent to the mobile terminal.

B) The SIP service proxy server in the network. The Service proxy server in the network can trigger customized user services. Generally, such services are executed on an application server. All mobile terminals that initiate a session establishment attempt must pass through the Service proxy server in the network. In this way, the proxy server can trigger the SIP service allocated to the user as appropriate. This means that some source hosts are required to ensure the correctness of these proxy servers. The visiting network can apply specific services and policies to introduce sessions. Therefore, the visiting network can contain a sip inbound proxy server to terminate the session. Generally, the SIP inbound proxy server is the same as the SIP outbound proxy server.

11) QoS requirements related to sip

A) QoS signaling and SIP independence

The selection of QoS signaling and resource allocation schemes must be independent of the selected session control protocol. This is considering the development of QoS Control and sip.

B) Coordination between SIP and QoS resource allocation

I) allocation before alerting. When creating a sip session, for an application, the resources required for request transmission establishment must be successfully allocated before the target user is notified. However, it should also be noted that it is also possible to notify users of SIP applications on a terminal before wireless resources are established.

Ii) the target user is added to the carrier negotiation. When a sip session is created, a final application allows the target user to join in to determine which carrier will be established. However, it is also possible to establish a sip session without user interference.

Iii) the carrier is successfully established. All end-to-end QoS signaling, negotiation, and resource allocation must be completed.

C) prevent service theft. Typically, if a user is assigned a QoS resource, there must be a permit control mechanism to prevent the user from exceeding the limits of network negotiation. The network must prevent unauthorized users from using unauthorized resources.

D) Authorize wireless resources. Because wireless resources are very expensive, networks must be able to manage them in one way. The Network must be able to identify who is using these resources and approve their use. For example, if the network does not supervise the use of wireless resources, an UA terminal can execute an unrestricted and uncontrolled resource reservation program.

E) prevent malicious use. 3gppims must prevent malicious use of the network by mobile devices. For example, a malicious UA cannot obey a program that involves the record-Route Header domain. When a concurrent request is sent, the UA can bypass the proxy server, which inserts a record-Route Header during initialization and processing.

F) prevent Denial of Service. The risk of DoS attacks received by a proxy server must be minimized. For example, a malicious ua can learn the IP address and port number of a sip Proxy Server (for example, the value in the record-Route Header) and establish an attack against this proxy server.

3. SIP Extension in IMS

3.1sip Compression

The session creation time may be extended due to the time required to transmit the SIP message through a limited bandwidth channel. Therefore, there must be a mechanism to effectively transmit sip signaling packets through wireless interfaces by compressing the SIP messages between UA and SIP exit proxy servers and between the SIP exit proxy server and UA. The IP address and the transport layer protocol header that sends the SIP messages must also be compressed.

1) compression and decompression of the SIP request and the response to the P-CSCF

The compression of SIP messages is an execution option. However, compression is strongly recommended. Because the compression support is mandatory, the UE can send or even be the first message to be compressed. Sigcomp provides a mechanism for the UE to know whether the State has been created in the P-CSCF.

UE must also extract SIP requests and response messages received from the P-CSCF. If ue detects a decompression failure in the P-CSCF, the repair mechanism will be executed and the algorithm can be changed.

The compression rules of SIP in the P-CSCF are the same as those of UE. The exchange of bytecode during registration will prevent unnecessary delays during session creation. The SIP request and the response to the UE also need to be compressed, and the response received from the UE also needs to be decompressed, they follow the same rules as the above P-CSCF.

2) compression operations are independent

The selected solution must be applicable to computation rules that cannot be compressed.

3) scalability of SIP Compression

The selected solution results must be scalable. When they are available, they are used in reverse compatibility to promote the merger of new and improved compression operations.

4) The minimal effect of SIP compression on the network

The impact of specific application compression on the existing 3GPP access network must be minimized. On the other hand, the compression mechanism must be independent of access. For example, compression must be defined on UA and exit sip proxy servers.

5) Availability of SIP Compression

It is necessary to make the use of sip signaling compression optional. To make it easier for mobile terminals to roam over a compressed network, mobile terminals must always support sip signaling compression. If compression is not supported, the communication can continue without compression, which depends on the local network's local policies.

6) compression Reliability

The compression mechanism should be reliable and can automatically fix Errors generated during decompression.

 3.2sip private Header

1) correlated uri (p-associated-Uri): transmits all associated Uris of mobile terminal registration addresses. It is used in the 200ok response to the Register request.

2) called Party identifier (p-called-party-ID): transfer the called identity. When a mobile terminal needs to register multiple Uris, the P-called-party-ID can be used to identify the real called Uri.

3) Visiting network identifier (p-visited-Network-ID): identifies the globally unique visiting network. Generally, a network identification is required for the scope of a P-CSCF, and the network identification code must be uniformly allocated by the network operator.

4) Access Network Information (p-access-Network-Info): transmits the wireless access technology and network information used by mobile terminals.

5) billing address (p-charging-function-address): IMS has two types of billing functions: the physical billing collection function (CCF) and the event billing function (ECF ). This message header field is used to indicate the physical address information of the billing function used. With this header field, IMS can achieve redundant storage of billing information.

6) billing vector (p-charging-vector): transmits the billing information in IMS, such as the billing collection point id, IP address, caller network identifier, and callee network identifier.

3.3 Security

The security authentication function of the mobile terminal is implemented by using the WWW-Authenticate and Authorization header fields of the SIP. When the UE sends a registration or call request to cscf, it must provide security parameters such as the Protocol identity and password in the Authorization header of the register message. When the UE does not contain security parameters, cscf sends a 401 response (unauthorized) to the UE, including the WWW-authenticate field. The www-authenticate field carries the necessary security parameters (such as the authentication method) for UE authentication ).

3.4 pre-processing precondition

In IMS, the availability of all necessary resources is the prerequisite for session establishment. Therefore, the introduction of SDP-based provision/response mechanism and related sip and sdpprecondition extensions are introduced. Precondition extended usage leads to a specific SIP call process. IMS controls media resources by using the go interface between ggsn and P-CSCF.

3.5 release a network-initiated call

In a mobile network, an ongoing call is required to be released because the signal is not covered or the battery is powered off. This problem can be solved by sending a bye request to ue from the network side. But this does not comply with the SIP principle, that is, the proxy server cannot send bye messages. However, due to the lack of a better solution, IETF accepts the requirements of 3GPP and this solution.

Parameters are extended for some sip headers. For example, parameters are extended for the WWW-Authenticate header, and a new auth-Param parameter field is defined, this field is used in the 401 (unauthorized) response to the Register request. This field also includes two specific parameters: integrityokey and cipher-key.

Added the "Application/pp-ims + XML" type for the message body MIME type in the SIP protocol, that is, ppip multimedia core subsystem extensible Language Body version 1, it is also agreed that this type of content cannot be sent out of the 3GPP network.

4. Typical sip process in IMS

The procedure for reaching the first Media Gateway Control Function (mgcf) to the IP Core Multimedia Subsystem initiated by the Public Telephone Exchange Network (PSTN) is as follows:

Step 1ss77 initial address message (IAM)

PSTN establishes a destination path to reach the media gateway (MGW) and initiates a signaling to the T signaling gateway (T-SGW) with the ss7iam message. Provides information about the identity and purpose of the relay.

Step 2 ipiam

T-SGW to connect SS7 messages, compress into an IP package to mgcf.

Step 3 interaction with H.248

Mgcf initiates an H.248 command to obtain the relay and IP port.

Step 4 invite (PSTN-O to S-S)

Mgcf initiated an invite request including an initial SDP, just like every unique S-CSCF to the S-CSCF process.

Step 5100 trying (S-S to PSTN-O)

Mgcf receives a 100trying temporary response, as specified by the S-CSCF-to-S-CSCF process.

Step 6183 sessionprocess (S-S to PSTN-O)

During each S-CSCF-to-S-CSCF process, the target performance of the media stream will be returned on a channel in the 183sessionprocess temporary response.

Step 7 prack (PSTN-O to S-S)

Mgcf decides the form of the final media stream for this session, and includes this information in the prack request, sent to each S-CSCF to the destination of the S-CSCF process.

Procedure 8200ok (S-S to PSTN-O)

The destination uses a 200ok message to respond to the prack request.

Step 9h.264 Interaction

Mgcf initiates an H.248 command to modify the connection parameters, indicating that MGW reserves the resources required for the session.

Step 10 reserve resources

MGW reserves resources required for sessions.

Step 11 comet (PSTN-O to S-S)

When the resource reservation is complete, mgcf sends a comet request to each S-CSCF to the terminal point of the S-CSCF process. SDP indicates that the resource is successfully reserved.

Steps 12200ok (S-S to PSTN-O)

The destination terminal uses a 200ok message to respond to the comet request.

Step 13180 ringing (S-S to PSTN-O)

The destination terminal can send signals freely. In this case, it uses a 180ringing temporary response to send a signal to the caller. This response is sent to each S-CSCF to the mgcf of the S-CSCF process.

Step 14 prack (PSTN-O to S-S)

Mgcf uses a prack request to respond to the 180ringing temporary response.

Step 15200ok (S-S to PSTN-O)

The destination terminal uses 200ok to respond to the prack request.

Step 16ip-acm

If the signal has been sent, mgcf will then send a full IP-ACM address to T-SGW.

Step 17acm

If the signal has been sent, the T-SGW continues to send an ss7acm message forward.

Steps 18200ok (S-S to PSTN-O)

When the call initiator responds, The S-S process will send a final response 200ok to mgcf to end.

Step 19ip-anm

Mgcf continues to send a IP-ANM Response Message to T-SGW.

Step 20 ANM

The T-SGW continues to send an ANM address full message to the PSTN.

Step 21. 248 Interaction

Mgcf initiates an H.248 command to change the MGW connection to make it bidirectional.

Step 22ack (PSTN-O to S-S)

Mgcf uses the ACK request to confirm the final response of 200ok.

5. Conclusion

Although the current IMS system architecture only supports mobile service access and does not support fixed access methods, the proposal of IMS conforms to the development trend of network convergence. Now there are ETSI/tispan (for fixed network applications), ITU-TFGNGN and many other standard entities involved. It is foreseeable by the three major trends of communication development (IP-based information organization, wireless information transmission, and multi-media information content). Driven by this development trend, h.323 and SIP will coexist for a considerable period of time, and because the SIP and IP methods are more compatible, with the gradual improvement of the SIP interoperability function, sip may eventually become a global protocol.