Cellular Network Infrastructure

h1. Voice Group Call (VGCS)

Voice Group Calls are part of the [[ASCI]] features introduced primarily for GSM-R.  They're basically the _bigger brother_ of [[Voice_Broadcast_Call]] (VBS).  So if you're unfamiliar with the former, it is suggested to first read up on VBS.

While VBS calls are unidirectional, VGCS calls are bi-directional between multiple parties.  There's always one _talker_ and many _listeners_.  The talker can change dynamically over time, basically creating a user experience of a classic two-way radio.  There is always only one talker active at any moment.

Like VBS, VGCS works with the concept of _groups_.  Eachs subscriber can be part of up to 50 groups, and calls are made to a specific group.  There can be any number of parallel groups active in parallel next to each other.  Each group has an _area_ describing a list of cells in which subscribers can participate in the group call.

* notification works via NCH, like VBS

* there still is only one TCH per BTS, like VBS

** however, that TCH has uplink: For voice, in case the talker is in the current cell

** for all listeners, the TCH is downlink-only, like in VBS

h2. Seizure of the uplink

* the NCH contains information if the uplink of a given group is currently free

** if it is, a subscriber wanting to talk can send an uplink request on the group TCH uplink

** if it is not free, the subscriber must request via RACH

h2. Arbitration

A subscriber who wishes to talk signals this to the network. The right to talk is allocated on a first-come-first-serve basis without queuing. 

h2. Specifications

For more information, see

* 3GPP TS 42.068 (Stage 1)

* 3GPP TS 43.068 (Stage 2)

* 3GPP TS 44.068 (Stage 3; GCC Layer 3 Protocol)