GSM Data Services

CSD follows the circuit-switching principle used by a voice call. When a CSD connection is established, GSM reserves a circuit between the mobile station and the interworking destination, such as an Internet Service Provider's modem bank. The resource remains assigned for the whole data call, including periods when the application is waiting for input. This gives predictable access to the bearer but leaves the radio resource occupied during silence in the application stream.

• Resource Allocation: The example connection uses one full-rate traffic channel (TCH/F), the same type of radio resource used for a voice call. Other bearer configurations can use different coding or channel options. The selected coding class changes the protection overhead and the user rate delivered to the interworking function.
• Billing Model: A circuit occupied a radio and core-network resource for the duration of the connection, so operators commonly billed CSD by connection time. The model charged for idle periods as well as transferred data and favoured short sessions with limited traffic.

A common CSD bearer provided $9.6 \text{ kbit/s}$ (kilobits per second). The value was roughly one fifth of a typical $56\text{k}$ fixed-line modem rate of that period. Bearer rate and coding options could differ, so 9.6 kbit/s is an example rather than a universal CSD value. Other configured bearers could use lower rates, such as 4.8 kbit/s, when additional protection was selected. The resulting capacity suited applications with small data volumes, such as:

• Sending and receiving plain-text emails on a connected laptop.
• Sending faxes from a portable fax machine or laptop.
• Early remote access for industrial or telemetry applications (e.g., a remote sensor sending back a small data reading).
• Basic text-only web browsing using the Wireless Application Protocol (WAP).

HSCSD extends CSD with a multislot mechanism. A network can assign several traffic-channel slots to one bearer, forming a higher-rate logical data connection. This allocation is often described as multislot allocation; the bearer parameters determine the resulting user rate. The terminal must maintain the assigned slots and the network must reserve them together, which limits the number of simultaneous multislot users in a cell.

Instead of assigning one TDMA time slot, an HSCSD network could allocate two, three, or four slots simultaneously to one data session. For a selected 9.6 kbit/s bearer, the arithmetic examples are:

• Two slots can provide $19.2 \text{ kbit/s}$ when each slot uses that 9.6 kbit/s bearer.
• Four slots can provide up to $38.4 \text{ kbit/s}$ under the same bearer assumption.

HSCSD increases the occupied radio resources in proportion to the allocated slots. Operators could therefore price it above single-slot CSD. A mobile transmitter also remains active during more positions in the TDMA frame, which can increase power consumption. HSCSD required compatible network and terminal support and later competed with packet-switched data services. The arithmetic describes radio bearer capacity; application throughput still depends on protocol headers, retransmissions, and the interworking network.

SMS uses signalling procedures instead of allocating a speech traffic channel for the complete message. A signalling channel carries the control messages and the short-message payload according to the selected radio procedure. The message protocol separates control information, destination addressing, validity information, and the user-data field, so a delivery attempt can be retried without repeating the original user action.

In classic circuit-switched GSM, SDCCH is a common path for SMS during call setup or idle-mode signalling, while other states and packet-domain procedures can use associated or packet signalling. The separation from a speech TCH provides these practical properties:

• A mobile station can exchange SMS during a voice call when the network supports the relevant associated-signalling procedure.
• The message uses signalling capacity and does not require a speech TCH for the complete payload transfer.

The Short Message Service Center (SMSC) stores and forwards messages. A simplified mobile-terminated flow is:

1. The mobile station submits the message to the operator's SMSC.
2. The SMSC obtains routing information for the destination through the mobile core, including the HLR function where applicable.
3. The SMSC forwards the message toward the serving MSC/VLR or another specified delivery node, which continues delivery to the mobile station.
4. If the destination is unavailable, the SMSC stores the message and retries according to its validity period, operator policy, and delivery procedures. A period such as 24 to 48 hours is an example configuration rather than a universal limit.

Store and forward separates message submission from final delivery. It supports later delivery after temporary unavailability, subject to validity, storage, routing, and error conditions.

A single SMS user-data field can contain 140 octets, so the bit budget is $140 * 8 = 1120$ bits. The GSM 7-bit default alphabet gives $1120 / 7 = 160$ septets in the basic single-segment case. Characters from the extension table consume an escape septet as well as the character septet. UCS2 uses 16 bits per character, giving $1120 / 16 = 70$ characters before other field rules are considered. An extension table can represent additional characters, and concatenated messages carry a user-data header, which reduces the payload available in each segment. The familiar 160 and 70 values therefore describe a basic single segment, not every encoded SMS.

• Calling Line Identification Presentation (CLIP) / Caller ID: The network presents the calling line identity to the called party when the subscription and privacy rules permit it. The identity is carried in call-control signalling for presentation by the terminal.
• Calling Line Identification Restriction (CLIR):CLIR requests restriction of the calling line identity at the called party. Subscription settings and the network's service-control rules determine whether the restriction applies.
• Call Waiting: The service notifies a subscriber about an incoming call during an active call. The subscriber can use call-control procedures to place the current call on hold, answer the waiting call, or reject it.
• Call Hold: The service suspends the user-plane speech path of an active call while retaining the call context. It can be used while another call is answered or before the original call is resumed.

This family of services redirects incoming calls to another destination, such as voicemail, an office phone, or another mobile number. GSM defines several forwarding conditions:

• Call Forwarding Unconditional (CFU): All incoming calls are redirected to the designated destination under the active subscription.
• Call Forwarding on Busy (CFB): Incoming calls are redirected when the served subscriber is busy according to the call-control state.
• Call Forwarding on No Reply (CFNRy): An incoming call is redirected after the configured no-reply timer expires. The timer value is service configuration, so 20 to 30 seconds is an example rather than a fixed GSM value.
• Call Forwarding on Not Reachable (CFNRc): An incoming call is redirected when the served mobile station cannot be reached, for example while it is switched off or outside coverage.

• Multi-Party Service (Conference Calling): The subscriber can join multiple incoming or outgoing call legs in one conference subject to the service and network capabilities.
• Barring of Calls: The subscriber or operator can restrict selected call classes, such as outgoing international calls or incoming calls while roaming, according to the active subscription.