From 26f7a7474e68a56f558700477c4ca38b77ce82a0 Mon Sep 17 00:00:00 2001 From: Philipp Maier Date: Thu, 23 Feb 2023 11:42:54 +0100 Subject: [PATCH] trau_pcu_ericsson: add comment about uplink blocks While testing each uplink block (CS1-C4, MCS1-MCS9) indvidually some specific behaviour was observed. Change-Id: I35b9a35c9b583378559daa400e105537338ee68a Related: OS#5198 --- include/osmocom/trau/trau_pcu_ericsson.h | 25 +++++++++++++++++++++++- 1 file changed, 24 insertions(+), 1 deletion(-) diff --git a/include/osmocom/trau/trau_pcu_ericsson.h b/include/osmocom/trau/trau_pcu_ericsson.h index 03a446f..37fa638 100644 --- a/include/osmocom/trau/trau_pcu_ericsson.h +++ b/include/osmocom/trau/trau_pcu_ericsson.h @@ -277,4 +277,27 @@ int er_gprs_trau_frame_decode_64k(struct er_gprs_trau_frame *fr, const ubit_t *b * value back. The timing adjustment is performed immediately within the same frame by appending or removing bits * at the end. This delays or advances the timing of the subsequent frame. * - * The procedure is very similar to the one described in: GSM 08.60, section 4.6.1.2 */ + * The procedure is very similar to the one described in: GSM 08.60, section 4.6.1.2 + * + * + * Usable coding schemes in uplink + * =============================== + * To understand which coding schemes are usable one must understand that the first CCU implementation for Ericsson + * RBS only supported CS1 and CS2 on 16K I.460 subslots. CS1 and CS2 were not supported due to the bandwidth limitation + * of the 16k subslots. In order to support higher bandwidth, in particular EDGE, a new CCU implementation was + * introduced. This newer implementation uses full 64K E1 timeslots and has full EDGE support MCS1-MCS9. + * + * Interestingly though, the newer CCU implementation seems to lack support for CS3-CS4. At least it was not possible + * To receive any CS3 or CS4 blocks from the CCU, while CS1 and CS2 work fine. In case the PCU is instructed to use + * CS3 or CS4 in downlink nothing is received. Since the CCU also demodulates noise and even sends the demodulation + * results back to the PCU there should have been at least some invalid CS3 and CS4 blocks in the demodulation results + * from time to time but also here no CS3 or CS4 blocks were observed. This leads to the vague assumption that the GPRS + * part of the CCU presumably never got updated and only EDGE support was added. + * + * It was also discovered that CS1 and CS2 only appear when the CCU is instructed to receive GMSK only + * (ER_UL_CHMOD_NB_GMSK). When the CCU is instructed to receive GMSK and PSK (ER_UL_CHMOD_NB_UNKN) CS2 does not appear. + * EGPRS related blocks (CS_OR_HDR_HDR1-3) appear in both situations. Since CS1 also plays a role in EGPRS it might be + * that the CCU is filtering non EDGE related blocks when ER_UL_CHMOD_NB_UNKN is used. + * + * The observations above were made with an RRUS 02 B3 + */