Preventive Cyclic Retransmission

8.5.1 Introduction

Preventive cyclic retransmission (PCR) is designed for use on signaling links with large propagation times (TP), for example, signaling links that are carried on satellite circuits (1.5.4). As in basic error correction, the FSN identifies the position of a MSU in its original sequence of transmission, and the BSN always identifies the most recently accepted message signal unit.

PCR uses positive acknowledgments only. Indicator bits FIB and BIB are ignored (they are permanently set to "1"), and incoming processing simply accepts or discards an error-free MSU based on the value of its FSN, which has to exceed the FSN of the most recently accepted MSU by one unit (modulo 128).

8.5.2 Preventive Retransmission Cycles

Whenever there is no LSSU or new MSU to be sent out, MTP2-A, in lieu of sending a FISU, starts a preventive retransmission cycle in which the MSUs in its retransmission buffer are retransmitted in-sequence and starting with the oldest one (lowest FSN). The retransmitted copies of MSUs that have already been accepted by MTP2-B now arrive out-of-sequence and are discarded. However, if any of the original MSUs were not accepted, their retransmission takes place much sooner than in basic error correction (where a time of at least 2TP is required for the negative acknowledgment to reach MTP2-A). This is why PCR can be used on signaling data links with propagation times that make basic error correction impractical.

A preventive retransmission cycle ends in one of two ways. If all MSUs in RB have been retransmitted and no LSSU or new MSU has to be sent out, MTP2-A starts sending FISUs. A retransmission cycle also ends when a LSSU or new MSU has to be sent.

0 0

Post a comment