- Service Provider Voice Implementation Models
- Residential Applications: Voice over Broadband
- Small/Medium Business Applications (Voice over T1/E1/PRI/CAS)
- IP Trunks
- Session Border Controller (SBC) Models
- Voice Security in Service Provider Networks
- Common Issues and Problems When Deploying IP-Based Telephony Services
- Summary
- References
Residential Applications: Voice over Broadband
In a voice over broadband deployment model, the SP uses the IP infrastructure to provide residential IP telephony services to its customers. An example of such an implementation model is the PacketCable architecture defined by Cable Television Laboratories (CableLabs) PacketCable specifications. The PacketCable specifications define a framework of how VoIP can be implemented over the Data Over Cable Service Interface Specification (DOCSIS)/IP infrastructure. Figure 3-2 provides a high-level overview of the PacketCable architecture. The system uses IP technology and QoS to provide high-quality transport for the VoIP network.
Figure 3-2 PacketCable Architecture Overview
The following are some of the key elements of the PacketCable network:
Call Management Server (CMS): The CMS is responsible for providing call control and signaling for the endpoints using Media Gateway Control Protocol/Network-Based Call Signaling (MGCP/NCS) protocol. In a centralized switched model, all the intelligence resides on the CMS, which is responsible for instructing other network elements on their functions.
The CMS is composed of several logical components, such as Gate Controller (GC), Media Gateway Controller (MGC), Signaling Gateway (SG), and Announcement Controller (ANC). The GC is responsible for quality of service (QoS) authorization and control. The MGC provides call control and signaling for PSTN Media Gateways. The SG communicates call signaling to the PSTN using protocols such as Signaling System 7 (SS7). The ANC interfaces with the Announcement Player (ANP) to play network announcements.
- Cable Modem Termination System (CMTS): The CMTS sits at the edge of the network and connects the endpoints to the SP infrastructure such as provisioning servers, CMS, Media Gateway (MGW), and so on over the DOCSIS Hybrid Fiber Coax (HFC) network. It also allocates resources for voice calls when instructed by the CMS and upon receiving requests from the endpoint.
- Media Terminal Adapter (MTA)/Embedded-MTA (EMTA): MTA connects the subscriber equipment, such as a host PC or analog phone, to the SP network over the DOCSIS (HFC) network. It establishes a physical connection with the CMTS and forwards traffic between the SP network and the subscriber equipment. It contains a network interface, radio frequency (RF) interface, CoderDecoder (CODEC), and all signaling and encapsulation functions required for VoIP transport, class features signaling, and QoS signaling.
- Media Gateways (MGW): The MGW provides bearer connectivity to the PSTN and is used for off-net calls (when an SP customer calls someone connected to the PSTN, basically an IP-to-PSTN network call).
- Provisioning Servers: Figure 3-2 includes a setup of servers; they perform provisioning and billing functionalities. These servers include the Dynamic Host Configuration Protocol (DHCP) server for assigning IP addresses and other network parameters to the endpoints, Domain Name Servers (DNS) for name resolution, Trivial File Transfer Protocol (TFTP) for downloading configuration files to MTAs, and optionally other servers such as syslog server and Ticket Granting Server (TGS), which are used in the PacketCable network.
- Application Servers: These servers include voicemail (VM) servers for providing voice mailbox service to subscribers, conferencing servers for audioconferencing service, announcement servers for playing network announcement messages, and Communications Assistance for Law Enforcement Act (CALEA) servers for subscriber wiretapping for law enforcement agencies.
- Record Keeping Server (RKS): These are used for billing purposes. They store call detail record information through PacketCable Event Messaging.
Residential gateways in the form of MTA embedded in a cable modem are also known as Embedded Multimedia Terminal Adapters (EMTA). VoIP access is provided at the customer premises. By plugging a standard analog telephone into the MTA device, a user can make phone calls to another Multiple System Operator (MSO) customer directly across the IP network or to anyone outside the SP or MSO network through an MGW.
CMSs and MGCs provide centralized call-control processing by passing control information and setting up connections between residential MTAs. After these connections are established, voice passes directly between gateway endpoints in the form of RTP packet streams, as shown in Figure 3-3. Most connections with the PSTN are through voice bearer trunks with a Media Gateway providing the bearer connections and a Signaling Gateway (SG) providing the signaling connection into the SS7 network. Multi-Frequency/Channel Associated Signaling (MF/CAS) trunks are provided for some specialized requirements, such as Operator Services.
Figure 3-3 PacketCable Signaling Architecture
The PacketCable Network-Based Call Signaling (NCS) protocol is used to communicate with the MTA endpoints. The PacketCable Trunking Gateway Call Signaling Protocol (TGCP) is used to communicate with Media Gateways. NCS and TGCP are profiles of the Multimedia Gateway Control Protocol (MGCP), which belongs to the xGCP suite of protocols. These protocols allow a central call control mechanism to control customer premises equipment (CPE) devices for voice services.