Introduction: Signaling is the name given to the act of sending status and command information to the network elements in a telecommunications network. The most basic commands are the on-hook / off-hook status information in a voice network. The most basic command information is call setup / tear down.

The current evolutionary state of signaling is different for TDM voice, for ATM, and for IP. The following sections cover the high-level basics.

Digital Voice Network Signaling: Although there are a large number of signaling variations in use today in the voice network, these can be categories as two basic types of signaling, in-band and out-of-band.

In terms of in-band signaling a major standard in use with TDM today is robbed bit signaling. Just as the name implies, some of the voice bits are robbed for signaling. In particular, the least significant bit of each voice sample in the sixth and twelfth (in D4 framing) and in the sixth, twelfth, eighteenth, and twenty-fourth frames (in ESF) are the A, B, C, and D signaling bits. The A, B, C, and D signaling bits convey whether the phone is on hook or off hook, and the dialed number. Robbed bit signaling is used mostly from PBXs to Telco switches, and from PBX to PBX in private networks.

Another form of in-band signaling is the TR303 signaling used in Subscriber Loop Carrier systems. This is used entirely in the loop plant of Telcos to obtain more than the standard 24 voice loop connections in the access plant on a single DS1. TR303 signaling is a hybrid of in-band and out of band signaling. Out of band signaling is used to convey off hook from the home to the local switch, and to set up a connection from the home to the local switch. After that in-band signaling is used for dialed digits, ringing, etc.

Between Telco switches the major standard in use in the world today is SS7 signaling. This is primarily out of band signaling, and is packet based. However, there is a difference in the European Standard than in the US Standard. In the United States, the signaling is quasi-associated, and is carried by separate facilities over separate paths between signaling end points. It may never traverse the same facilities as the actual voice path. In the European Standard, the signaling is associated, and rides on designated, reserved channels in the same facility that carries the voice channels.

SS7 signaling has a layered stack protocol, with three major layers. The lowest layer in the stack is the Message Transfer Part (MTP). MTP handles all communications between the signaling end points. This includes ensuring that the communications are of reasonable quality, taking action when appropriate and alarming when it is not. The next layer in the stack is the ISDN User Part (ISUP). ISUP consists of the commands for basic call setup and takedown. ISUP messages can be loosely sorted onto Call Control, Maintenance, and Network Administration. For example Call Control includes the negotiation the switches have to complete to assign the resources for the transaction. The last layer in the stack is called Transaction Capability Application Part (TCAP). TCAP handles the transfer of information from network databases to switches. The transferred information pertains to the type of service, where to route the call (i.e. time of day routing), where to bill the call, calling card information, whether the call is to be admitted to the network (fraud control), etc. TCAP is required for some IN services and all AIN services.

SS7 signaling endpoints consist of Switches, Signal Transfer Points (STPs), Service Control Points (SCPs, a.k.a. NCPs). STPs are used to route the signaling packets from Switch to Switch and Switch to SCP. In essence STPs are packet switches, and do not use the actual signaling information to process a voice call. SCPs are databases that contain service and customer specific information on how a call should be handled. This includes routing, billing, and other service features.

The design of an SS7 signaling network is a complex exercise, involving capacity management, reliability management, restoration, and quality management. It involves understanding and managing the many inter-company interfaces that a modern signaling network has. The design must be robust enough to handle many rare and sometimes unforeseen events, because if the signaling network is disrupted, no calls can be processed, and billing for in progress calls could be lost.

ATM Signaling: ATM signaling does not have as rich a set of commands as SS7. ATM signaling currently consists of Broadband ISDN User Part (BISUP), which is an extension of ISUP. ATM has the call setup routing commands, however it does not yet have the TCAP database query capability that narrowband SS7 has for voice traffic and services. Even so the call setup and routing capability currently available with ATM does provide a powerful signaling capability for a core-switching vehicle.

The addressing scheme in ATM consists of a 2-part address, the Virtual Path Identifier (VPI) and the Virtual Circuit Identifier (VCI). VPIs and VCIs have to be defined between switches. The VPI can be considered to be a trunk group like entity, while the VCI can be considered to be a trunk like entity. A successful BISUP negotiation between switches result in the assignment of resources between the switches in terms of a VPI/VCI for a specific customer use, until no longer needed.

IP Signaling: IP signaling, as with ATM, consists only of call control at this point in time. All signaling in IP is carried in the header, and centers on address resolution. The process of address resolution, i.e. going from IP address to physical address is a significant problem that has been solved.

In order to effect address resolution, IP address formats consist of 3 sub-fields in the 32 bit address. If the first bit is a 0, then the address is in a Class A network, the next 7 bits specify the network identifier, and the next 24 bits specify the host identifier. If the first 2 bits are 10, then the address is in a Class B network, the next 14 bits specify the network identifier, and the next 16 bits specify the host identifier. If the first three bits are 110, then the address is in a Class C network, the next 21 bits specify the network identifier, and the next 8 bits specify the host identifier. If the first 4 bits are 1110, then the address is in a Class D network, and the next 28 bits specify a multicast address.