Sunteți pe pagina 1din 13

IMS, IP Multimedia Subsystem Tutorial

- a summary or tutorial of IMS, IP Multimedia Subsystem, providing the

highlights of its operation and structure.
IMS, or IP Multimedia Subsystem is having a major impact on the telecommunications industry, both
wired and wire-less.
Although IMS was originally created for mobile applications by 3GPP and 3GPP2, its use is more
widespread as fixed line providers are also being forced to find ways of integrating mobile or mobile
associated technologies into their portfolios.
As a result the use of IMS, IP multimedia subsystem is crossing the frontiers of mobile, wire-less and
fixed line technologies. Indeed there is very little within IMS that is wireless or mobile specific, and as
a result there are no barriers to its use in any telecommunications environment.

IMS basics
IMS, IP multimedia subsystem, itself is not a technology, but rather it is an architecture. It is based
on Internet standards which are currently the major way to deliver services on new networks.
However one of the key enablers for the architecture is the Session Initiation Protocol (SIP), a
protocol that has been devised for establishing, managing and terminating sessions on IP networks.
The overall IMS architecture uses a number of components to enable multimedia based sessions
between two or more end devices.
One of the elements is a presence server that handles the user status, and this is a key element for
applications such as Push to talk over Cellular (PoC) where the presence, or user status is key to
enabling one user to be able to talk to another.
With users now needing to activate many sessions using different applications and often
concurrently, IMS provides a common IP interface so that signalling, traffic, and application
development are greatly simplified. In addition to this an IMS architecture means that subscribers
can connect to a network using multiple mobile and fixed devices and technologies. With a variety of
new applications from Push to talk over Cellular (PoC), gaming, video and more becoming available,
it will be necessary to be able to integrate them seamlessly for users to be able to gain the most
from these new applications.
It also has advantages for operators as well. Apart from enabling them to maximise their revenues,
functions including billing, and "access approval" can be unified across the applications on the
network, thereby considerably simplifying this area.

IMS development & history

IMS was developed by the cellular industry but to meet the growing needs across the mobile, fixed
and IT / computing networks.

It was developed out of a need for the telecommunication industry, and in particular the cellular
telecommunications industry to be able to allow for ubiquitous access to multimedia services from
any terminal.
IMS grew out of the political landscape of the day. This shaped many elements of its design and
architecture, and as a result, it needs to viewed with this in mind.
The IMS standards were developed by a group called 3G.IP which was formed in 1999. This group
was soon taken under 3GPP where its work could be better harmonised with the work of the cellular
industry who it appeared would be the main users.
Accordingly IMS is defined within the 3GPP standards and its development can be tracked within the
different releases.





First introduction of IMS



IMS emergency services

Combinational services
Voice call continuity



Single radio voice call continuity (SR-VCC)

Multimedia telephony



IMS centralised services

IMS continuity services
Multimedia interworking between IMS and CS networks
IMS multimedia Telephony and supplementary services



IMS emergency calls over GPRS and enhanced packet

system, EPS
Enhancements of IMS customised alerting tone service
IMS restoration services



IMS services continuity - inter-device transfer


One major push for its use came at Mobiel World Congress 2010, where GSMA announced they
were supporting what was then termed the "One World" initiative for carrying Voice over LTE, VoLTE.
As the system was based around the use of IMS, many operators then decided it was necessary to
incorporate IMS capabilities within their networks.

IMS Architecture
- IMS architecture giving overview of structure and details of the elements.
The IMS architecture is relatively complicated and this can mean it is expensive to implement and
also requires attention to understand.
The IMS architecture can be split down to make it more accessible to understand.
The IMS architecture consists of many different entities which may be collocated or distributed within
the network.

IMS architecture basics

The architecture of an IMS system can be split into a number of main elements or areas:

User equipment:

As the name implies, the user equipment or UE is part of the IMS

architecture resides with the user - it is the endpoint.

Access network:

This is the portion of the IMS architecture through which the overall

network is accessed.

Core network: This is a major element within the IMS architecture and provides all the core

Application layer:

The application layer contains the web portal and the application

servers, which provide the end user with service and enhanced service controls. T

IMS architecture functional view

Although a complete architecture diagram is quite complicated, a general overview provides a more
informative view.

Simplified view of the functional IMS architecture

The overall IMS architecture contains a number of main elements:

IMS CSCF - Call Session Control Server:

The IMS CSCF is the section of the

architecture that provides the registration of the endpoints. It also provides routing for the SIP
signalling messages. It also links to the interworking and transport layer to provide QoS. The

Server CSCF: : This element in the overall IMS CSCF is a session control entity for
endpoint devices and it maintains session state.

Proxy CSCF: : This part of the IMS CSCF is the entry point to IMS for devices. The
P-CSCF is the first point of contact for the UE and it forwards SIP messages to the
user's home S-CSCF. It provides device control interworking security. Within the PCSCF, the PCF or Policy Control Function provides QoS management.

Interrogating CSCF: : This entity within the IMS CSCF is a session control entity for
endpoint devices that maintains session state.

Home Subscriber Server, HSS: This is an important element within the IMS architecture
which provides the subscriber data base for the home network.

Breakout gateway control function, BGCF: This entity within the IMS architecture selects
the network in which a PSTN breakout is to occur. If this is to occur in the same network as
the BGCF, then the BGCF selects a media gateway control function, MGCF

Media gateway control function, MGCF:

This entity interworks the SIP signalling. It

manages the distribution of sessions across multiple media gateways.

Media server function control, MSCF:

This manages the use of resources on media


SIP applications server, SIP-AS: The SIP-AS is a service execution platform on which one
or more services are deployed.

IMS core network

As the name implies the IMS core network is at the centre of the network, and accommodates some
of the main features within the network as a whole. Typically the entities within the core network
address the functions with its basic operation.
Some of the main entities within the core network include:





Much of the core network may be co-located, or alternatively it may be distributed around the
Although the path between entities located within the same premises may be shorter and occur
more quickly, there are reasons for multiple deployment of the same entities around the network.

Network capacity is a key reason for multiple deployment. With load increasing it can be
advantageous to have several instances of the same entity to cope with peak demand.
Multiple HSSs may be needed to hold details of all the network subscribers, or multiple SCSCFs may be needed to handle the peak number of SIP sessions that may occur.

Geographic distribution can provide real benefits if the network is distributed over a wide
area. Although the system would operate if centred on a small area, geographic distribution
can reduce the traffic as fewer nodes require to be accessed and latency can also be

Distributing the network over a wide area can add redundancy for instances when one centre
may experience a power outage, etc. It is possible, for example for one S-CSCF to take over
the user's registration dynamically from another, or HSS data may need to be accessed from
different occurrences of the HSS. This approach adds significant resilience to the network
and considerably increases the reliability.

IMS access network

The IMS access network is made up of those elements that are associated with communication from
the core network to the outside world - external networks and users.
The IMS network can be accessed through various forms of IP Carrier Access Networks, IP-CAN.
The IP-CAN provides the IP connectivity as well as mobility. The IMS terminal sends control plane
signalling and media transfer through the IP-CAN to the IMS core network.

IMS HSS Home Subscriber Server

- IMS HSS - Home Subscriber Server details, structure, and operation within
the overall IMS architecture.
The IMS HSS or home subscriber server is the main subscriber database used within IMS.

The IMS HSS provides details of the subscribers to the other entities within the IMS network,
enabling users to be granted access or not dependent upon their status.

IMS HSS basics

When a subscriber registers onto an IMS network, the subscription data is retrieved from the HSS by
the Serving-CSCF, S-CSCF that has been assigned to the subscriber.

IMS S-CSCF Serving Call State Control

- details for the function, operation and implementation of the IMS S-CSCF,
Serving Call State Control Function.
The IMS S-CSCF, or Serving Call State Control Function, is one of the major entities within the
overall IMS architecture.
The IMS S-CSCF is at the core of many functions within an IMS network, communicating with many
other functions within the overall system.
The S-CSCF undertakes a variety of actions within the overall system, and it has a number of
interfaces to enable it to communicate with other entities within the overall system.

IMS S-SCSF basics

The Serving Call State Control Function, S-CSCF, is the main SIP session control node within the
overall IMS network.
When a subscriber enters the network, the subscriber provides a contact address and a public user
identity. This is provided to the S-CSCF.
These two elements of data are then linked in a process known as binding.
In view of its role within the IMS network, the S-CSCF is seen as the registrar for the network,
although the HSS holds the data against which the S-CSCF checks the authenticity of the subscriber
requesting entry.

S-CSCF interfaces
The S-CSCF interfaces with several other entities within the IMS network.

IMS S-CSCF Interfaces

Simplified view of the IMS architecture showing IP-CAN

The actual registration process has various stages to it. The primary communications occur between
the HSS and the S-CSCF, but other IMS entities are also involved.
The following data transfers occur:

Subscription data is transferred from the HSS to the S-CSCF. This includes items such as
the IMS public user identity (public identifiers used for identifying he subscriber for originating
and terminating media sessions), and the service trigger data.

Charging subscription data is sent from the HSS to the P-CSCF via the S-CSCF.

IMS public user identity data which is also kept in the S-CSCF is forwarded to the user

Service subscription data is forwarded from the HSS to the SIP-AS via the S-CSCF. This
enables charging for the services to be managed.

IMS P-CSCF Proxy Call State Control

- details for the function, operation and implementation of the IMS P-CSCF,
Proxy Call State Control Function.
The P-CSCF is the user to network proxy. In this respect all SIP signalling to and from the user runs
via the P-CSCF whether in the home or a visited network.

P-CSCF basics
When any user registers with the IMS network, the registration signalling will pass through the PCSCF.
For instances where a subscriber has more than one active terminal, these may communicate and
pass registration requests through different P-CSCFs, but the same S-CSCF will always be used.
This is because the HSS will provide data to use the same S-CSCF regardless of which I-CSCF
presents the requests.

IMS I-CSCF Interrogating Call State Control

- details for the function, operation and implementation of the IMS I-CSCF,
Interrogating Call State Control Function.
The I-CSCF, Interrogating Call State Control Function is one of the main elements within the overall
IMS hardware architecture.
The I-CSCF is used for forwarding an initial SIP request to the S-CSCF. When the initiator does not
know which S-CSCF should receive the request.

I-CSCF basics
The I-CSCF, Interrogating Call State Control Function is a key element in the IMS roaming
methodology. It enables requests to be routed to the correct Serving Call State Control Function. As
there may be several S-CSCFs either within a network, or if a roaming user requests access.

The I-CSCF interrogates the HSS to obtain the address of the relevant S-CSCF to process the SIP
initiation request.
The SIP request is routed via the I-CSCF to the S-CSCF using the following stages:









P-CSCF forwards registration request to I-CSCF

I-CSCF enquires from HSS which S-CSCF should receive the SIP message and
handle data

During SIP session establishment: There are again a number of transactions involved:

SIP request sent to I-CSCF

I-CSCF contacts HSS to ascertain which S-CSCF should receive the SIP message

Standalone SIP transaction:

The process is the same as for the SIP session

It is possible that when the I-CSCF has interrogated the HSS and no S-CSCF has been assigned.
Under these circumstances the HSS provides an S-CSCF capabilities description to the I-CSCF.
The I-CSCF may then assign a suitable S-CSCF and then forward the SIP request to that S-CSCF to
be actioned.

IMS Layers & Stack

- details of the OSI layers as used within the IMS stack, including the
Transport & Endpoint, Session Control, & Application Layers.
When referring to communication between systems a layered model is often used. The OSI, Open
System Interconnection reference model is widely used.
This seven layer OSI stack is used as the basis for many systems, and the same is true for IMS.
While the OSI layer system is a generalised system, it is easily adapted for use with IMS, or any
other system.

OSI layers and IMS

The OSI layer reference model was developed by the International Organisation for Standardisation,
ISO. It is a model that is used in communications systems to divide the communication channel into
various levels of tasks.

OSI layer reference model and IMS

Each layer provides some distinct and well defined services to the adjacent layer further up the
However, above layer 5 (and sometimes layer 4) the distinction can become a little less defined, and
some services overlap the layers. Often layer 5 and above may become more specific to the
environment, e.g. IMS in which they are used. As a result, many systems will have their own
terminology and particular definitions for these layers.

IMS layers & stack basics

There are three main IMS layers that are tailored for this application. They correspond to layers 4
and above. These are:

Transport and Endpoint Layer [equates to OSI layer 4]

Session Control Layer [equates to OSI layer 5]

Application Server Layer [equates to OSI layers 6 & 7]

IMS Transport and Endpoint Layer

This IMS layer initiates and terminates the SIP signalling, setting up sessions and providing bearer
services including the conversion from analogue or digital formats to packets.
This IMS layer also contains all of the media processing facilities including media gateways. These
can be used to convert VoIP bearer streams to the PSTN TDM format. They can also be used to
provide many media-related services such as conferencing, playing announcements, collecting inband signalling tones, speech recognition, and speech synthesis.

IMS Session Control Layer

This layer contains what is termed the Call Session Control Function (CSCF) which provides the
endpoints for the registration and routing for the SIP signalling messages, enabling them to be
routed to the correct application servers. The CSCF also enables QoS to be guaranteed. It achieves
this by communicating with the transport and endpoint layer.
The layer also includes other elements including the Home Subscriber Server (HSS) that maintains
the user profiles including their registration details as well as preferences and the like. It includes the
presence server essential to many interactive applications such as PoC. A further element of the
session Control Layer is the Media Gateway Control.

Application Server Layer

The control of the end services required by the user is undertaken by the Application Server Layer.
The IMS architecture and SIP signalling has been designed to be flexible and in this way it is
possible to support a variety of telephony and non-telephony servers concurrently.
Within this layer there is a wide variety of different servers that are supported. This includes a
Telephony Application Server (TAS), IP Multimedia - Service Switching Function (IM-SSF),
Supplemental Telephony Application Server, Non-Telephony Application Server, Open Service
Access - Gateway (OSA-GW), etc.

SIP - Session Initiation Protocol

- overview or tutorial about the basics of SIP, the Session Initiation Protocol
used in a variety of telecommunication applications including VoIP, Voice over
SIP, the Session Initiation Protocol is used in many applications and has been adopted as the
signalling protocol for use with Voice over IP ( VoIP ). SIP is a signalling protocol that is used for
establishing sessions on an IP network. The presence of SIP enables sessions to be set up in a way
that enables a host of new services to be made available, thereby allowing far greater flexibility to be
SIP, Session Initiation Protocol, is focussed purely on establishing, modifying and terminating
sessions, and has no interest in the content of the sessions. In view of the focus of SIP, it provides a
level of simplicity that enables to be extensible, and to site easily within different deployment
architectures and scenarios.
SIP is an RFC standard - RFC 3261 from the Internet Engineering Task Force (IETF). This is the
organization that is responsible for administering and developing the mechanisms that support the
Internet. While other protocols have been used in the past, SIP has now become the protocol of
choice as a result of its flexibility and ability to be updated.

Key functions
There are a number of key functions that SIP provides. It is able to provide name translation and
user location, it negotiates the features that will be available in a session and it manages the
participants in a session.

User location and name translation - this function enables data to reach a party regardless of
location. To achieve this SIP, Session Initiation Protocol addresses are used. These are very
similar in format to email addresses, having elements such as a domain name and a user
name or phone number. Also because of their structure, they are easy to associate with
email addresses.

Feature negotiation - as different parties may have different features that are supported it is
necessary that both ends communicate in a way that both can support. For example it would
be no use a video enabled phone trying to sent video to a voice only phone. Thus when a
link is set up all participants negotiate to agree the features that are supported. Also when
one user leaves a session, the remaining ones may renegotiate to determine whether any
new features may be supported.

Participant management - sessions need to be managed to enable users to enter or leave

sessions. SIP provides this capability.

SIP elements
SIP comprises two basic elements, namely the SIP User Agent and the SIP Network Server:

The SIP User Agent This is the component of the protocol that resides with the user. In turn it
consists of two parts: the User Agent Client (UAC) which initiates the calls and the User
Agent Server (UAS) which answers calls. It allows calls to be made using a peer to peer
client server protocol.

SIP network server This element contains three basic parts: the SIP Stateful Server, the SIP
Stateless Server, and thirdly the SIP Redirect Server. These servers act to provide the
location of the user and accordingly direct data to the user, and they also provide name
resolution in a similar way that email addresses and domain names do on the Internet as it is
unlikely that users will remember IP addresses.

SIP also provides its own transfer mechanism which is independent of the packet layer. This enables
it to perform reliably over protocols such as UDP - a particularly useful feature under some