S1 Handover Preparation Failure

Seeing lots of S1 handover - Handover Preparation failure. Core is from Baicells and EnB are mixture of Baicells, Huawei, ZTE. T304 timer is at max 2000ms.

We have a single core.

What other issues can cause this?

correct definition of many cm stuff, like s1 ip route correctly set, neighbor cell information correctly configured, source cell correctly configured in target cell, also MME with accurate infomation, if you have pcap share so we can have more background.

1 Like

Attach is a sample PCAP file.
HO Preparation Failure sample sample.zip (121.3 KB)

Ok Bro so far I saw the following, theres only S1AP related to handover messages.
10.55.88.35 β†’ mme
you have the following type of ues in your capture
uetype1 β†’ rel9/cat4 ->band 4,13
uetype2 β†’ rel9/cat3 ->band 2,4,5,17
uetype3 β†’ rel12/cat4 ->1,2,3,5,8,13,17
uetype4 β†’ rel9/cat3 ->band 2,4,5,13,17,25

the euarfcns used are 5255(13) β†’ 700 MHz , 2536(5) β†’ 850 MHz

The following are the ips of the enodebs used:
image

The ones in yellow are β€˜co-site’ cells.

This is the distribution of the cause for the S1AP messages
image

majority of messages are related to handover(preparation, notification, cancelation, etc)

So for the moment I will recommend to doble check the definitions in the MME in terms of tac,enbid,lcid
Also if possible take a look at mme logs.
Now if nothing of this works, then take a look at the FGI in some of the messages and doble check in your mme configuration to see if a feature is supported or not.
FGI looks like this:

For example the UE used doesnt support handover from FDD to TDD, and there are many more features you can check in terms of compatibility.
Also it would be worth to take a look at any configuration of the S1-U like ips etc.
Hope your case is an easy one and not a hard one, good luck!

1 Like

Thanks for the thorough analysis. We don’t have to configure the EnodeB on the core (MME). They simply register with the MME. I reviewed the neighbor list several times. TAC, EnBID and PCI settings are all fine. I will see what EnodeB logs we can generate to figure out exactly why the MME returns HO Preparation failure. I could not find an exact reason in the tcap traces.

Ok, so its time to go for the compatibility reported by the ue and what do you have available in the mme.

Most of our UEs are the same. 95% of our customers use portable MIFI. We see them handover on other EnB. WE see successful HO from same UE type on the same EnB that are reporting HO Prep failure.

Can you do IMSI trace on the ue having trouble? this also will help.

I will get a TEMS or NEMO license/phone and try to get some traces in the field. I have access to MME logs.