CDR Processing¶

Overview¶

Call Detail Record (CDR) processing is a critical component of the offline charging system. This document describes the CDR processing pipeline and best practices for handling charging data records.

CDR Types¶

Voice CDRs¶

• Mobile Originated Calls (MOC)
• Mobile Terminated Calls (MTC)
• Roaming Calls
• Supplementary Services

Data CDRs¶

• GPRS/Data Usage
• Packet Data Protocol (PDP) Context
• Serving Gateway Records
• PDN Gateway Records

Messaging CDRs¶

• SMS Records
• MMS Records
• Messaging Gateway Records

Processing Pipeline¶

interface CDRProcessingStage {
  id: string;
  name: string;
  status: 'PENDING' | 'PROCESSING' | 'COMPLETED' | 'FAILED';
  startTime: string;
  endTime?: string;
  error?: string;
  metadata: Record<string, unknown>;
}

interface CDRBatch {
  id: string;
  source: string;
  recordCount: number;
  fileSize: number;
  checksum: string;
  stages: CDRProcessingStage[];
  createdAt: string;
  completedAt?: string;
}

Stages¶

1. Collection
2. Receive raw CDR files
3. Validate file integrity

4. Store in staging area

5. Validation

6. Schema validation
7. Mandatory field checks
8. Data type verification

9. Duplicate detection

10. Enrichment

11. Add subscriber information
12. Resolve reference data
13. Apply business rules

14. Calculate derived fields

15. Aggregation

16. Group related records
17. Calculate summaries
18. Apply rating rules

19. Generate invoicing data

20. Storage

21. Archive raw CDRs
22. Store processed records
23. Update billing database
24. Generate reports

Error Handling¶

enum CDRError {
  INVALID_FORMAT = 'INVALID_FORMAT',
  MISSING_FIELDS = 'MISSING_FIELDS',
  DUPLICATE_RECORD = 'DUPLICATE_RECORD',
  PROCESSING_ERROR = 'PROCESSING_ERROR',
  STORAGE_ERROR = 'STORAGE_ERROR'
}

interface CDRErrorReport {
  batchId: string;
  recordId?: string;
  error: CDRError;
  message: string;
  timestamp: string;
  retryCount: number;
  rawData?: string;
}

Implementation¶

File Processing¶

interface CDRFile {
  id: string;
  filename: string;
  format: 'ASN1' | 'CSV' | 'XML' | 'JSON';
  size: number;
  recordCount: number;
  checksum: string;
  status: 'NEW' | 'PROCESSING' | 'COMPLETED' | 'FAILED';
  createdAt: string;
  processedAt?: string;
}

interface CDRProcessor {
  validateFile(file: CDRFile): Promise<boolean>;
  processRecords(file: CDRFile): Promise<CDRBatch>;
  handleErrors(errors: CDRErrorReport[]): Promise<void>;
  generateReports(batch: CDRBatch): Promise<void>;
}

Performance Optimization¶

1. Batch Processing
2. Optimal batch sizes
3. Parallel processing

4. Resource management

5. Caching

6. Reference data caching
7. Rate plan caching

8. Subscriber info caching

9. Storage

10. Efficient indexing
11. Partitioning strategy
12. Archive management

Monitoring¶

Key metrics to track: - Processing throughput - Error rates - Processing latency - Resource utilization - Queue lengths

Best Practices¶

1. Data Integrity
2. Implement checksums
3. Validate record counts

4. Maintain audit trails

5. Error Recovery

6. Implement retry logic
7. Store failed records

8. Alert on failures

9. Performance

10. Monitor system load
11. Optimize queries
12. Scale horizontally

References¶

1. 3GPP TS 32.297: Charging Data Record (CDR) file format and transfer
2. 3GPP TS 32.298: Charging Data Record (CDR) parameter description
3. 3GPP TS 32.240: Charging architecture and principles