Public access mobile radio systems (PAMR) using trunking technology allow more efficient use of the frequency spectrum, as all the users of the system jointly use a common pool of channels. Trunked PAMR systems can operate on two different levels of management, on the basis of messages (message trunking) and transmissions (transmission trunking). When using message trunking, the system allocates the radio channel from the beginning to the end of the conversation. However, with transmission trunking, the system allocates a radio channel when one partner presses his or her push-to-talk button (PTT) and deallocates the radio channel when he or she releases the PTT button to change talking partners. Thus, during the silence produced by this change, no channel is occupied. The advantage of transmission trunking is that there is no waste of channel occupancy during the pauses in the conversation. The drawback is that a more complex control system with more signalling is needed, where the distribution of the transmission occupancy is more crucial in the performance than in the distribution of the message holding time when trunking by message. In message trunking, the lack of radio channel only causes a delay and the user has to wait to begin the conversation, while in transmission trunking, it can produce an unpleasant effect of clipped words. This paper verifies that voice calls, generally assumed to have an exponential holding time, are better modelled with other distributions, and analyses the statistical properties of the conversation time in two different PAMR systems (13 refs.)

Peer Reviewed