LTE Handover Fault Diagnosis

LTE Handover Fault Diagnosis

Confidential Information of Huawei. No Spreading Without Permission












Inter-frequency handover and inter-!" inter-!" handover can #e coverage-#ased$ coverage-#ased$ load-#ased or service-#ased handover. "his course focuses on coverage-#ased handover.



!%&!'&!( event can #e used to trigger coverage-#ased inter-frequency handover. eNode) can configure which event is used.



)*&)+ event can #e used to trigger coverage-#ased inter-!" inter-!" handover. eNode) can configure which event is used.





!* leaving condition ：,s  Hys  "hresh



!+ leaving condition：,s / Hys 0 "hresh



!% leaving condition：,n  1fn  1cn  Hys  ,s  1fs  1cs  1ff



!' leaving condition ：,n  1fn  1cn  Hys  "hresh



!( leaving condition ：,s - Hys 0 "hresh* or ,n  1fn  1cn  Hys  "hresh+



)* leaving condition ：,n  1fn  Hys  "hresh



)+ leaving condition ：,s - Hys 0 "hresh* or ,n  1fn  1cn  Hys  "hresh+



"he varia#les in the formula are defined as follows2 











Ms is the measurement result of the serving cell$ not ta3ing into account any offsets. Mn is the measurement result of the neigh#ouring cell$ not ta3ing into account any offsets. Ofn is the frequency specific offset of the frequency of the neigh#our cell. Ocn is the cell specific offset of the neigh#our$ and set to 4ero if not configured for the neigh#our cell. Ofs is the frequency specific offset of the serving frequency. Ocs is the cell specific offset of the serving$ and is set to 4ero if not configured for the serving cell.



Hys is the hysteresis parameter for the events.



Thresh1 and Thresh2 is the threshold parameter for the events.





Currently intra-frequency handover is most popular in networ3s. So intra-frequency handover parameters are discussed here.







Step *2 56 measures source cell and target cells. If the trigger cond ition is fulfilled$ 56 reports measurement results.



Step + and step %2 !fter receiving measurement result$ eNode) ma3es decision to trigger handover. Source eNode) sends Handover Request message to target eNode)$ including target cell I7 and some parameters for handover preparation. !fter receiving handover request$ target eNode) tries access control and prepares resource for handover$ then sends Handover Request Acknowledge to source eNode).



Step '2 Source eNode) sends RRC Connection Reconfiguration including handover parameters for handover.



Step (2 Source eNode) sends S Status Transfer message to target eNode).



Step 82 56 triggers non-contention #ased random access in target cell.



Step 92 !fter successful random access in target cell$ 56 sends RRC Connection Reconfiguration Co!"lete message to target eNode).



Step : and step ;2 "arget eNode) sends #ath Switch message to ,,6. ,,6 informs S


Step *=2 "arget eNode) sends $% Conte&t Release message to source eNode) to release the resource of the 56.







,easurement control includes measurement I7$ measurement quantity$ measurement o#>ect$ measurement report configuration and etc.



"he figure in the slide is an e?ample of measurement o#>ect.





,easurement report configuration contains report configuration I7$ measurement quantity$ report mode and etc.





,easurement report contains measurement I7 and measurement result. 

,easurement I7 in measurement report is same with the I7 in the corresponding measurement control.



"he following is SP and S@ measurement report mapping2 RSRP measurement report mapping Reported value RSRP_00 RSRP_01 RSRP_02 … RSRP_95 RSRP_96 RSRP_97

Measured quantity value RSRP < -140 -140 ≤ RSRP < -139 -139 ≤ RSRP < -138 … -46 ≤ RSRP < -45 -45 ≤ RSRP < -44 -44 ≤ RSRP

Unit dBm dBm dBm … dBm dBm dBm

RSRQ measurement report mapping Reported value RSRQ_00 RSRQ_01 RSRQ_02 … RSRQ_32 RSRQ_33 RSRQ_34

Measured quantity value RSRQ < -19.5 -19.5 ≤ RSRQ < -19 -19 ≤ RSRQ < -18.5 … -4 ≤ RSRQ < -3.5 -3.5 ≤ RSRQ < -3 -3 ≤ RSRQ

Unit dB dB dB … dB dB dB





Handover command contains PCI of target cell and some related configuration information.







If whole-networ3 handover API is not good$ first transmission and equipment alarms should #e chec3ed$ then cell-level handover API is analy4ed.





B.HH1.IntraeN).Intrareq.6?ecSucc1ut：Num#er of successful intra-eNode) intrafrequency outgoing handovers in a cell



B.HH1.IntraeN).Interreq.6?ecSucc1ut：Num#er of successful intra-eNode) interfrequency outgoing handovers in a cell



B.HH1.IntereN).Intrareq.6?ecSucc1ut：Num#er of successful inter-eNode) intrafrequency outgoing handovers in a cell



B.HH1.IntereN).Interreq.6?ecSucc1ut：Num#er of successful inter-eNode) interfrequency outgoing handovers in a cell



B.HH1.IntraeN).Intrareq.6?ec!tt1ut：Num#er of intra-eNode) intra-frequency outgoing handover attempts in a cell



B.HH1.IntraeN).Interreq.6?ec!tt1ut：Num#er of intra-eNode) inter-frequency outgoing handover attempts in a cell



B.HH1.IntereN).Intrareq.6?ec!tt1ut：Num#er of inter-eNode) intra-frequency outgoing handover attempts in a cell



B.HH1.IntereN).Interreq.6?ec!tt1ut：Num#er of inter-eNode) inter-frequency outgoing handover attempts in a cell









"he result is from a networ3 with H5!W6I equipments.









Possi#le scenarios for handover triggered too early2 

!fter receiving handover command$ 56 fails to handover to the target cell #ecause of the #ad signal in target cell$ then 56 triggers C connection reesta#lishment in source cell.



56 handovers to target cell successfully #ut downlin3 synchroni4ation fails in the target cell$ then 56 triggers C connection reesta#lishment.







Possi#le scenarios for handover triggered too late2 

"he source cell signal is #ad$ so 56 can not receive handover command. "hen 56 triggers C connection reesta#lishment in target cell. If the 56 conte?t is setup successfully in the target cell$ the C reesta#lishment can #e done.



"he source cell signal drops too fast$ 56 has no time to report measurement result. "hen 56 triggers C connection reesta#lishment in target cell. "he 56 conte?t can not #e setup successfully in the target cell$ the C reesta#lishment can not #e done.



"here is something wrong with the D+ transmission$ HA'O(%R R%)$%ST message can not #e transferred to the target cell.











In real networ3s$ most handover failures are caused #y 5u interface pro#lems$ resulting in the a#normal signaling over 5u interface.







"hese are >ust possi#le reasons for a#normal 5u signaling. ,ore information and trace result should #e collected for deep analysis.





SP2 SP is the downlin3 reference signal power received #y 56. SP and SIN roughly reflect the downlin3 quality.


SIN2 SIN is the downlin3 reference signal SIN. If S IN is less than =d)$ the downlin3 channel quality is poor. If SIN is less than -%d)$ the downlin3 channel quality is very poor and close to the demodulation threshold$ which can easily cause loss of handover messages. "he uplin3 SIN can #e o#tained from the user performance tracing on the B,".



I)B62


5plin3 scheduling grants and downlin3 scheduling assignments over the P7CCHs2 It is the num#er of P7CCHs correctly demodulated #y a 56 in each ""I. If there is enough uplin3 or downlin3 data source and the channel quality is very good$ the nu m#er of the grants or assignments could #e very high$ even close to *===. If the num#er of uplin3 grants is too low$ it may #e caused #y #ad channel quality.





or e?ample$ poor downlin3 channel quality not only affects demodulation of the downlin3 signaling$ the incorrectly demodulated P7CCH also affects the uplin3 scheduling$ leading to loss of uplin3 signaling.







!#normal signaling over D+ interface is usually caused #y D+ transmission pro#lem or internal pro#lem of site. Internal pro#lem of site is not popular. So it is very possi#le that D+ transmission causes a#normal signaling over D+ interface. It is not easy to analy4e transmission pro#lem. In this slide transmission pro#lem is not the focus.









Possi#le solutions2 

!d>usting the antenna to change the coverage of serving cell or neigh#oring cell



!d>usting CI1 to ma3e handover happen earlier 

"o increase neigh#oring cell CI1 or decrease serving cell CI1 cou ld ma3e handover happen earlier.





!2 and !ccess esponse



In handover area$ SP is good #ut SIN and I)B6 is #ad. So it is very possi#le that the pro#lem is caused #y downlin3 interference.





56 can not receive random access response #ecause of the downlin3 interference.













"he figure shows the SP$ SIN and I)B6 during handover. I)B6 and SIN is poor$ so handover fails.





"he figures shows that after ad>ustment handover is triggered successfully. SIN and I)B6 are #etter.














LTE Handover Fault Diagnosis

Recommend Documents