The IMS specifications define functions to handle the signaling and subscriber traffic for multimedia applications. The functions are separated into logical layers, and many of the specified functions often reside in a single platform. Vendors have the flexibility to implement IMS functions in consolidated ways, and it is natural that platforms such as softswitches will combine many logically separate IMS call-processing functions, and that routers will take on some of the session-enforcement and gateway functionality in IMS.
The three layers are the service layer, the control layer, and the connectivity layer. Figure 1 shows a high-level view of the IMS architecture.
Service layer—Hosts application and content services, including application servers and Web servers. It also includes generic service enablers that manage service elements such as user groups and presence. These service elements connect to subscribers through the control plane. The application layer supports most of the multimedia applications or application enablers, such as presence and location of the subscriber.
Control layer—Makes the policy decisions that are enforced in the connectivity layer. This layer provides session control and management, and is responsible for setting up and taking down packet sessions. It also contains information about subscriber authentication, service authorization, and location.
Connectivity layer—Supports the core network architecture of the General Packet Radio Service (GPRS), which consists of support nodes for data services. This layer is where routers, switches, firewalls, and optical transport reside, along with gateways that translate protocols between packet- and circuit-based traffic.
Session Initiation Protocol (SIP) is the main signaling protocol in IMS. SIP is the proposed standard for multimedia communication between subscribers interacting with voice, video, and instant messaging. In IMS, the use of SIP facilitates interconnectivity between fixed and mobile networks.