Ask the Expert!

Forum Post: Authentication Failure with Cause "Non EPS authentication unacceptable"

Sumit — Fri, 01 Mar 2013 13:48:00 GMT

Hi All,

During Attach if UE sends Authentication Failure with Cause "Non EPS authentication unacceptable". What will MME do?

a) Sends Identity Request to get IMSI and then map with GUTI. If GUTI Maps correctly. MME proceed with Attach Request

b) Sends Authentication Request again

c) Sends Authenthentication Reject if Attach is with IMSI

Also this Case is Applicable in Normal Attach or Combined Attach. According to me it should be Combined Attach.

Forum Post: RE: Inactivity timer in LTE

Yongren — Fri, 01 Mar 2013 10:05:00 GMT

Does "inactivity timer is implementation dependent" mean network does not need to implement the S1 release procedure by inactivity timer in eNB which is described in TS23.401 s5.3.5?

Forum Post: LTE NAS Protocol

SUBHAM SINGH — Fri, 01 Mar 2013 09:37:00 GMT

What is the significance of KSI (Key Set Identifier) bits in lte.

KSI is of 2 types- either KSI-asme or KSI-SGSN.

It is represented in 3 bits.The purpose if KSI bits is to reuse of base key K-ASME. What is the purpose of these 3 bits of KSI?

Why it is 3 bits?

Forum Post: RE: Reg: How RLC can distinguish MIB, SIB messages in TM Mode

decent — Fri, 01 Mar 2013 04:56:00 GMT

Hi Expert,

I am not getting the clarity on how RLC can map to MAC about the MIB, SIB's, & other SI's using the only 1 BCCH Channel or while implementing we have to use separate BCCH channels for above So that MAC can recognize SIB1, SIB2......& MIB & Other SI;s . Or Other way what I got is map the RRC directly to MAC without RLC So that there wont be any mapping issues. But I know that last scenario is not the correct method. So Please clarify my doubt.

Forum Post: RE: RRC connection release

Naveen Jain — Fri, 01 Mar 2013 03:21:00 GMT

Do both keep their default bearer?

Forum Post: RE: TB size vs PRB size

Biolog — Fri, 01 Mar 2013 02:55:00 GMT

The terminology in the LTE specs is confusing.

To clarify the confusion, 1 PRB (normal cyclic prefix) = 12 subcarriers * 7 OFDM symbols = 84 Resource Elements (not 168 RE).

However, the N_PRB parameter used in 3GPP TS 36.213, Table 7.1.7.2.1-1 refers to a pair of PRB's,

not just one PRB. (See 3GPP TS 36.211, 6.2.3.)

It specifically states:

- A virtual resource block is of the same size as a physical resource block. (VRB = PRB in size.)

- A pair of virtual resource blocks over two slots in a subframe is assigned together by a

single virtual resource block number, N_VRB. (N_PRB = N_VRB since VRB = PRB.)

Hence, when the N_PRB parameter is 1 for instance, 2 PRB's is implied and

hence 84 RE * 2 PRB = 168 RE.

Forum Post: RE: C-RNTI

Biolog — Fri, 01 Mar 2013 01:40:00 GMT

I understand the need to differentiate the TPC C-RNTI versus the other C-RNTI identities since

TPC C-RNTI uses DCI Format 3 & 3A which has ambiguous payload size that is identical to DCI

Format 0 & 1A.

But for the other C-RNTI identities (Temporary C-RNTI, C-RNTI, & SPS C-RNTI), they all contain

unambiguous DCI payload lengths and hence I failed to see the need for differentiation.

The only benefit I can see is the reduction in the # of blind decoding attempts being made on the

PDCCH to identify the correct DCI format used.

Would this be the only reason for the existence of the other C-RNTI identies?

Or is there a more prevailing reason?

Forum Post: Different MCS index for one UE

orpa — Fri, 01 Mar 2013 00:17:00 GMT

Hi expert,

In some paper regarding adaptive modulation and coding, I read that different modulation and coding level can be applied to different subcarrier depending on the application requirement. In video transmission base layer bitstream and enhancement layer bit stream can be transmitted using different modulation and coding. My question is that is it possible in LTE to use different Modulation and coding in different resource block for one UE.

Forum Post: RE: LTE FACH channel

Don — Thu, 28 Feb 2013 23:07:00 GMT

I'm nott aware of a FACH channel in LTE; I've only run across it in UMTS. Are you thinking of the MCH channel for MBMS?

Forum Post: RE: why maximum supported size of a PDCP SDU is 8188 octets?

Don — Thu, 28 Feb 2013 22:57:00 GMT

1500 bytes is a typical MTU for IPv4 packet, but IPv6 packets can be much larger. The LTE system is designed to handle very large packets, especially as the Internet moves towards IPv6.

Forum Post: RE: LTE signal as a passive Radar

Don — Thu, 28 Feb 2013 22:55:00 GMT

That depends on what you mean by passive radar. As far as I'm aware, there's no spec for such a thing, and in general, wireless communications systems do not operate in bands that are suitable for radar; they're designed to penetrate obstacles, not reflect from them.

In FDD LTE systems, there is no guard time between frames.

Forum Post: RE: Excessive paging Vs resource utilization

Don — Thu, 28 Feb 2013 22:45:00 GMT

Paging messages occupy some number of RBs, along with some of the PDCCH capacity, so yes, a lot of paging can adversely affect the net throughput over the air. Depending on the MME and eNB vendor's implementation, they may throttle the total amount of paging performed in a frame, and discard excess paging messages.

20 paging messages per second doesn't strike me as a lot, though, relative to the total capacity of the LTE channel. It's annoying and I hope you track down the culprit, but the rest of the subscribers probably won't notice the impact.

Forum Post: RE: A3 Event Parameters and HO Optimization

sohail.bhatti — Thu, 28 Feb 2013 22:41:00 GMT

Thanks Ovidiu

Forum Post: RE: Multiple preparation

Don — Thu, 28 Feb 2013 22:37:00 GMT

Multiple preparation allows the source eNB to simultaneously request handover resources from multiple target eNBs, even though only one would be used. This minimizes delays if the preferred target fails for some reason. As far as I'm aware, this is not a standardized LTE capability (although there's nothing preventing this from happening); it's a vendor-specific implementation feature.

Forum Post: RE: LTE QoS Classification

Don — Thu, 28 Feb 2013 22:29:00 GMT

That would depend on the type of services your VIP customer would be using. For Internet applications, QCI 6 is the highest-priority "general" bearer, QCI 8 and 9 being the others, which would give the user preference over the general popoulation. I wouldn't recommend using a GBR bearer for data services, since that would kill your network capacity.

Don't use QCI 5 for data services. It's intended specifically for IMS services, in order to ensure that call control will be successfully even under congested situations.

In general, non-GBR bearers are not affected by admission control, at least as far as resource and scheduler reservation is concerned; they would still be subject to total bearer limits, though.

Forum Post: RE: How long does UE issue RRCConnectionSetupComplete after it receives RRCConnectionSetup?

Don — Thu, 28 Feb 2013 22:13:00 GMT

The eNB will generally run its own internal timer waiting for the RRCConnectionSetupComplete message, typically on the order of a couple of seconds; it's not listed in the specs since the timer value doesn't need to be exchanged with any other entity.

I'm not sure what resources you're concerned about here. Uplink resources assigned to the UE to send the RRCConnectionSetupComplete message are only availble for that specific subframe, and are not reserved after that. If the eNB doesn't detect the Complete message, it will usually send the Setup message again, in case it was lost on the downlink. If there is still no response, it will discard the RRC context and peg an access failure.

Forum Post: RE: Reference signal positions in the RBs

Don — Thu, 28 Feb 2013 22:02:00 GMT

In spec 36.211, section 6.10.1.2, the mapping of reference signals to resource elements is based on, in part, the cell's physical cell ID (NcellID).

Forum Post: RE: Create Session Response with Cause "UE context without TFT already activated"

Don — Thu, 28 Feb 2013 21:59:00 GMT

The specs aren't entirely clear how this error code is generated, but the equivalent error in 24.301 (NAS Signaling), error code #46 "EPS Bearer Context without TFT already activated" is defined as "This ESM cause is used by the network or the UE to indicate that it has already activated an EPS bearer context without TFT." Basically, it implies that the system is attempting to establish a session or bearer that is already active, probably due to a lost response message that timed out. Note that default bearers initially don't have TFTs associated with them, although they can be defined later.

Forum Post: RE: MO Signalling vs MO Data

Don — Thu, 28 Feb 2013 21:45:00 GMT

The basic difference is that MO Signaling sessions are intended to exchange signaling messages only with the network (such as Tracking Area Updates), while MO Data sessions involve the exchange of user data (such as emails or web browsing).

Forum Post: any reference or good pdf about HARQ and rate matching in lte ????

mai fouad — Wed, 27 Feb 2013 22:45:00 GMT

please i want any reference or good pdf about HARQ and rate matching in LTE (detailed explanation)???