LTE / 5G NR
Radio Network Optimization
Interview Prep

📋 Accessibility Interview Questions

📋 Retainability Interview Questions

📋 Mobility Interview Questions

📋 Integrity Interview Questions

🧱 The QCI / Bearer Model — know this cold

🎛️ The Optimization Levers

📊 VoLTE KPIs & Targets

🛠️ Step-by-Step Optimization Workflow

🔍 Troubleshooting Playbook

🎧 MOS Scale (know the numbers)

📋 VoLTE Interview Questions

-- Basic KPI: CSSR by cell and hour
WITH raw AS (
  SELECT
    cell_id,
    date_hour,
    SUM(rrc_conn_estab_succ)  AS rrc_succ,
    SUM(rrc_conn_estab_att)   AS rrc_att,
    SUM(erab_estab_init_succ) AS erab_succ,
    SUM(erab_estab_init_att)  AS erab_att
  FROM mr_kpi_hourly
  WHERE dt BETWEEN '2025-06-01' AND '2025-06-07'
  GROUP BY 1, 2
)
SELECT
  cell_id,
  date_hour,
  ROUND(rrc_succ * 100.0 / NULLIF(rrc_att, 0), 2)  AS rrc_sr_pct,
  ROUND(erab_succ * 100.0 / NULLIF(erab_att, 0), 2) AS erab_sr_pct,
  ROUND(rrc_succ * 100.0 / NULLIF(rrc_att,0)
      * erab_succ / NULLIF(erab_att,0), 2)  AS cssr_pct
FROM raw
ORDER BY cssr_pct ASC
LIMIT 50;

-- Worst cells by DCR with window rank
SELECT
  cell_id,
  site_id,
  ROUND(erab_rel_abnormal * 100.0 / NULLIF(erab_rel_total, 0), 2) AS dcr,
  RANK() OVER (ORDER BY
    erab_rel_abnormal * 1.0 / NULLIF(erab_rel_total, 0) DESC) AS rank
FROM mr_kpi_daily
WHERE dt = '2025-06-07'
HAVING erab_rel_total > 100   -- filter low-traffic cells
ORDER BY rank
LIMIT 20;

📋 SQL Interview Questions

# KPI calculation from raw counters CSV
import pandas as pd

df = pd.read_csv('kpi_raw.csv')

# Compute CSSR and DCR
df['rrc_sr']  = df['rrc_succ']  / df['rrc_att'].replace(0, pd.NA) * 100
df['erab_sr'] = df['erab_succ'] / df['erab_att'].replace(0, pd.NA) * 100
df['cssr']    = df['rrc_sr'] * df['erab_sr'] / 100
df['dcr']     = df['erab_rel_abn'] / df['erab_rel_tot'].replace(0, pd.NA) * 100

# Flag underperforming cells
df['bad_cssr'] = df['cssr'] < 99.0
df['bad_dcr']  = df['dcr']  > 0.5

worst = df[df['bad_cssr'] | df['bad_dcr']].sort_values('cssr')
print(worst[['cell_id', 'cssr', 'dcr']].head(20))

# Log parsing — extract RLF events from eNB trace
import re
from datetime import datetime

pattern = re.compile(r'(\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}).*RLF.*cell=(\w+).*cause=(\w+)')
events = []

with open('enb_trace.log') as f:
    for line in f:
        m = pattern.search(line)
        if m:
            events.append({'ts': m.group(1), 'cell': m.group(2), 'cause': m.group(3)})

df_rlf = pd.DataFrame(events)
print(df_rlf.groupby(['cell', 'cause']).size().reset_index(name='count').sort_values('count', ascending=False))

📋 Python Interview Questions