Case Study: High NR NSA Drop Rate on Top 11 Cells
Problem:
In a live NR NSA network, a consistently high drop rate was observed across the top 11 NR cells. The issue was persistent and impacting overall user experience, particularly for data and VoLTE services.
Analysis:
Initial checks included reviewing alarms, OPTLOGS, and BRD/CHR logs for both NR and LTE layers. No service-impacting alarms or recent configuration changes were identified. Further KPI and counter analysis highlighted that the major contributors to the drop rate were SgNB abnormal releases and UE lost events. This was confirmed through high values in counters such as N.NsaDc.SgNB.AbnormRel.Radio and N.NsaDc.SgNB.AbnormRel.Radio.UeLost. The findings indicated that uplink-related issues, specifically poor uplink data reception and decoding, were leading to frequent connection drops.
Solution:
To resolve the issue, the parameter ABNUEULDATAERRORJUDGETHLD was modified from 0 to 5 across all affected NR cells using the command:MOD GNBOAMPARAM: ABNUEULDATAERRORJUDGETHLD=5;
This parameter defines the uplink error rate threshold for identifying abnormal UEs. A value of 0 disables abnormal UE detection, while a non-zero value enables the system to identify UEs experiencing high uplink errors. After applying the change, the NR drop rate improved significantly.
Recommendation:
It is recommended to configure ABNUEULDATAERRORJUDGETHLD with an optimized non-zero value (typically 3–10) to ensure proper abnormal UE detection, improve uplink reliability, and reduce unnecessary drops in NSA networks.
Case Study: 5G NR SgNB Drop Rate Degradation at Network Level
Problem:
Following the launch of 5G NR service, the network experienced a consistently high 5G NR SgNB Drop Rate across the network. The degradation was observed from the first day of service launch and impacted overall NSA mobility performance and user experience.
Analysis:
Initial optimization efforts focused on the top worst-performing sites from a coverage perspective; however, no significant KPI improvement was achieved. This indicated that the issue was not primarily related to RF coverage. Further investigation shifted toward transport and inter-node connectivity analysis, where X2 interface performance and SON-related configurations were reviewed. It was identified that the X2 SON feature was either not properly enabled or not optimally configured across multiple sites, directly impacting SgNB stability and dual connectivity retention.
Root Cause:
Detailed parameter audit revealed three major issues contributing to the degraded SgNB drop performance. First, the X2 SON Delete Timer for X2 Fault was configured with excessively high values, delaying fault clearance and causing stale X2 relationships to persist. Second, the X2 Setup by SON switch was not enabled on all sites, preventing automatic establishment and recovery of X2 neighbor relations. Third, several X2 SON parameters were found to be improperly configured, resulting in inefficient X2 setup and fault handling behavior.
Solution:
To resolve the issue, X2 SON functionality was enabled and optimized across the affected network. The X2 SON Delete Timer for X2 Fault was reduced from 10080 minutes to 2880 minutes at gNB level to allow faster removal of faulty X2 links. At eNB level, the X2 Setup by SON switch was enabled on all relevant sites, and X2 SON setup/failure thresholds and timers were standardized according to optimization guidelines. Additional X2 SON setup policy parameters were also corrected to ensure proper automatic X2 neighbor establishment and maintenance.
Case Study: NR NSA High Drop Rate After 5G New Site Integration
Problem:
After integrating new 5G sites into the network, it was observed that the EN-DC SCG Drop Rate was significantly high and failed to meet the defined acceptance threshold. The issue was impacting NSA service stability and degrading user experience on newly integrated sites.
Analysis:
Initial troubleshooting included checking site availability, reviewing active alarms, and analyzing board logs and CHR traces for both LTE and NR layers. No major hardware or transport issues were identified. Further investigation revealed that the Abnormal SgNB Release Rate, including MeNB-triggered abnormal releases, was elevated at approximately 20–30%. Detailed CHR analysis confirmed that the dominant failure reason was T310 timer expiry, indicating radio link failure during NSA operation. It was also observed that a portion of connected UEs did not support 70 MHz NR bandwidth, yet SCG addition was still attempted, causing unnecessary SCG failures.
Root Cause:
The high EN-DC SCG drop rate was caused by multiple NSA radio and configuration-related factors. The primary contributor was T310 expiry, resulting in abnormal SgNB releases during weak or unstable NR radio conditions. In addition, SCG failures were aggravated by delayed release behavior and bearer inactivity timer settings. Another key issue was that SCG addition was being attempted for UEs that did not support 70 MHz NR bandwidth, leading to unsuccessful SCG setup attempts and increased failure rates.
Solution:
A multi-step optimization plan was implemented across the affected sites. First, the T310 optimization switch (SCG_REL_MEAS_OPT_SW) was enabled to improve T310 expiry handling and reduce unnecessary abnormal releases. Second, SCG failure handling was optimized by adjusting inactivity and re-establishment related parameters, including RrcRelDelayTmr, BEARERACTIVITYTHLD, and enabling enhanced SCG failure full configuration recovery. Third, NR bandwidth capability check optimization (NR_SCG_BW_CHECK_SW) was enabled on the LTE side to prevent SCG addition for UEs that do not support 70 MHz NR cells. Additional abnormal uplink UE detection enhancements and bearer activity threshold tuning were also implemented to improve release decision accuracy.
Result:
Following implementation of the optimization actions, significant KPI improvement was achieved. On the top optimized sites, EN-DC SCG Drop Rate reduced from 7.45% to 0.35%, while EN-DC Setup Success Rate improved from 93.67% to 99.92%. The optimization successfully stabilized NSA dual connectivity performance and brought the affected sites within acceptance thresholds.
Recommendation:
This solution is recommended for deployment across 80+ NR sites with 70 MHz bandwidth throughout the network to proactively reduce SCG drops, improve EN-DC setup success, and enhance overall NSA user experience.