Dear Anas,

1/ this is an interesting question, as this is not clearly stated in the standard. In my undestanding, and according to the tests I made with comercial devices (Quectel BC95), I can say yes: UE has time to at least send NPRACH, wait for random access response. If the RAR contains its RAPID, the attachment prodecure goes on, else the next NPRACH attempt happens.

2/ This is not described in 3GPP standard because this is left to the eNB manufacturer. Many strategies can be considered at eNB side: fixed backoff, adapted backoff. So my response is: every strategy is suitable as soon as it is not in contradiction with 3GPP standard.

I hope this helps,

Best regards,

Vincent

Thank you very much Vincent Savaux

Your answer was very helpful

Though from my understanding of question 1: Msg3, Msg4, and data transmission should happen all before the next periodicity.

However, we have some insight from 3GPP standards

In fact, mac-Contention-ResolutionTimer which is the window for receiving Msg4 transmission can be between pp1 to pp64 in which pp indicates NPDCCH period. But what is the actual period of pp in milliseconds?

I know that it varies according to the number of repetitions but what is the relation between pp and repetitions in time. example: for 1 repetition, what is the actual time of pp?

Though from my understanding of question 1: Msg3, Msg4, and data transmission should happen all before the next periodicity.

-> I don't think so: the periodicity only indicates when the UE is allowed to transmit NPRACH. It is independent of the other scheduling parameters for Msg3 and Msg4.

In fact, mac-Contention-ResolutionTimer which is the window for receiving Msg4 transmission can be between pp1 to pp64 in which pp indicates NPDCCH period. But what is the actual period of pp in milliseconds?

-> I never met pp in the standard, do you have a reference please?

Best regards,

Vincent

Thank you very much Vincent Savaux

I'm sorry for the late reply as I was checking the information you mentioned in the standards and some papers

-> I thought that periodicity is like a timer or slot (slotted-Aloha) system that works for all UEs simultaneously. However, if, as you are saying, periodicity is independent of Msg3, Msg4 and other scheduling parameters, does that mean that when eNB gives a grant of Msg3 for a certain UE, all other UEs will pause their periodicity timer until that UE finishes its transmission completely?

Or does it mean that each UE has its own periodicity timer? but then there would be collisions between some UEs who are doing Rach and others who are scheduling

-> pp is the time interval between two successive NPDCCH

it is mentioned several times as a measure of time in TS 136 331 which can be found in the following link

https://www.etsi.org/deliver/etsi_ts/136300_136399/136331/15.03.00_60/ts_136331v150300p.pdf

for example: on page 783 in the table for ra-ResponseWindowSize-r13 they mentioned pp2, pp3, pp4, pp5, pp6, pp7, pp8, pp10

Thanks again!

Anas

Dear Anas,

does that mean that when eNB gives a grant of Msg3 for a certain UE, all other UEs will pause their periodicity timer until that UE finishes its transmission completely?

-> yes and no: first, keep in mind that NPRACH transmission happens rarely, only when UEs wake up, or when it is required by the eNB. Once a UE is attached to an eNB, it should not transmit NPRACH. Second, eNB transmits NPRACH parameters through SIB2, and 3 sets of parameters can be used (3 CE levels), according to the signal quality (linked to received signal power). Then, all UEs in the cell do not transmit NPRACH with the same power, repetitions number, periodicity, etc. Given

a CE level, if some UEs transmit NPRACH in the same time, which is not likely, then the eNB transmits grant to one UE, and only the UE that are expecting this random access response (RAR) will receive this Msg3. The choosen UE keeps go on the random access procedure, the other ones wait for the next NPRACH window, indicated by the NPRACH periodicity, to transmit NPRACH again.

Or does it mean that each UE has its own periodicity timer? but then there would be collisions between some UEs who are doing Rach and others who are scheduling

-> As in the first response indicated, no: periodicity is triggered by UES according to: 1/ 3 sets of parameters transmitted in SIB2, 2/ among the 3 sets, UE choose one according to the received signal strength.

Collisions may happen, but it is not likely, since NPRACH transmission is rare. If it happens, the random access procedure includes contention resolution.

-> pp is the time interval between two successive NPDCCH

-> OK, this indicates the window within which UEs expect a RAR from eNB

I hope my resposne is clear,

Best regards,

Vincent