Call Detail Records (CDRs) serve as a fundamental building block for tracking, billing, and analyzing various communication services like voice calls and SMS. These records are generated by different network elements such as the Mobile Switching Center (MSC) and the Intelligent Network (IN). However, discrepancies between these records can lead to serious issues, including revenue leakage, fraud, and data inconsistencies. This blog post explores why CDR reconciliation between MSC and IN is crucial, how the process can be implemented for both voice (MO) and SMS (MO) services, the challenges encountered, and how Big Data tools like Apache Spark can significantly enhance the efficiency and accuracy of this process.

Why CDR Reconciliation is Important?

CDR reconciliation is not just a technical exercise; it is a crucial component of an effective revenue assurance process. Ensuring that all CDRs from different network nodes match perfectly is essential for several reasons. From maximizing revenue capture to ensuring compliance, a reliable reconciliation process can prevent numerous issues that telecom operators face on a daily basis. Let’s delve into the key reasons why this process is indispensable:

• Revenue Assurance: Reconciliation ensures that all billable events are captured and billed correctly. Discrepancies between MSC and IN CDRs can lead to revenue leakage, where services are provided but not billed accurately.
• Financial Integrity: Reconciliation validates the consistency of call records across different network elements, ensuring that the data used for billing, analysis, and reporting is accurate, thereby safeguarding financial integrity.
• Operational Efficiency: Identifying missing or duplicate records during reconciliation helps in detecting errors in the network, such as incomplete calls or unbilled events, which could lead to potential revenue loss and operational inefficiencies.
• Regulatory Compliance: Telecom operators must often report accurate traffic data to regulatory authorities. CDR reconciliation helps in ensuring the accuracy of these reports, thereby meeting regulatory requirements and avoiding penalties.

Achieving CDR Reconciliation for Voice (MO) and SMS (MO)

Implementing an effective CDR reconciliation process requires a systematic approach to match records from MSC and IN. While the goal is to ensure that both sets of records align perfectly, the actual process involves several steps and considerations, especially when dealing with multiple fields like msisdn, imsi, caller, called, timestamp, and sometimes a unique correlation ID. Here’s how you can achieve a robust reconciliation process for voice and SMS services:

Reconciliation Process for Voice (MO) between MSC and IN:

• Identify Key Fields: The main fields to consider for reconciliation are msisdn, imsi, caller, called, timestamp, and a unique correlation ID.
• Intermediate CDRs Aggregation: MSC often generates intermediate CDRs for the same call, especially for long-duration calls that are split into smaller segments. These intermediate CDRs must be aggregated based on caller, called, and timestamp to create a complete call record before reconciling with IN CDRs.
• Unique Correlation ID: Ideally, there should be a unique identifier that links the CDRs from MSC and IN for the same call. This ID ensures that the call records can be matched accurately across the two systems.
• Timestamp Matching: In cases where a unique correlation ID is not available, we can use a combination of fields like caller, called, and timestamp to identify matching records. However, time differences between MSC and IN systems can pose challenges.

Reconciliation Process for SMS (MO) between MSC and IN:

• Identify Key Fields: Similar to voice, important fields for SMS reconciliation include msisdn, imsi, sender, receiver, timestamp, and a unique correlation ID (if available).
• Aggregation of Long SMS Messages: For SMS, particularly long messages that are split into multiple segments (e.g., concatenated SMS), it is important to aggregate these segments into a single CDR based on sender, receiver, and timestamp before proceeding with reconciliation. This ensures that the entire message is accounted for correctly, avoiding partial billing or data discrepancies.
• Matching Criteria: After aggregation, it is crucial to match the sender and receiver numbers along with the timestamp to accurately reconcile the SMS records between MSC and IN.

Challenges in CDR Reconciliation

Despite the critical need for CDR reconciliation, the process is often challenging. From the absence of a unique correlation ID to time discrepancies between different network elements, telecom operators often face significant obstacles in achieving accurate and consistent CDR reconciliation. Understanding these challenges is essential to developing strategies to mitigate them. Here are some of the most common issues faced during the reconciliation process:

• Lack of Unique Correlation ID: Many times, there is no unique correlation ID between MSC and IN, making it difficult to directly match the records. In such cases, using a combination of msisdn, caller, called, and timestamp becomes necessary.
• Intermediate CDRs for Voice Calls: For long-duration calls, MSC generates multiple intermediate CDRs. These need to be aggregated to form a single complete call record before any reconciliation can be performed with IN CDRs. Failure to do so can result in mismatches and inaccuracies.
• Segmented SMS Messages: Long SMS messages that are split into multiple segments must be aggregated before reconciliation. Missing or incomplete segments can lead to discrepancies between MSC and IN records.
• Time Difference Between MSC and IN: Even a slight time difference between the CDR generation in MSC and IN can cause mismatches. For instance, if the MSC and IN systems are not synchronized to the same time source, the same event may appear at slightly different times, complicating the reconciliation process.
• Handling Duplicate Records: Sometimes, multiple CDRs are generated for the same event, especially in cases of call setup failures or retries. These duplicates must be filtered out to avoid false discrepancies.
• High Data Volume: The large volume of CDRs generated daily in a telecom network can make reconciliation a resource-intensive and time-consuming process.

How Big Data Tools Like Apache Spark Can Help

With the rise of Big Data technologies, the telecom industry now has powerful tools to handle large-scale data reconciliation more efficiently. Apache Spark, in particular, stands out for its ability to process massive datasets in parallel, making it an excellent choice for CDR reconciliation. Leveraging Spark can not only speed up the reconciliation process but also provide additional capabilities like anomaly detection and real-time analytics. Here’s how Spark can transform the reconciliation process:

• Scalability and Speed: Spark can handle large volumes of data efficiently with its in-memory processing capabilities, making it ideal for processing millions of CDRs quickly.
• Distributed Processing: Spark’s distributed architecture allows it to parallelize the reconciliation process across multiple nodes, significantly speeding up the comparison of records from MSC and IN.
• Data Transformation: Spark provides powerful APIs for data manipulation, making it easier to preprocess the CDR data (e.g., filtering out duplicates, handling missing values, aligning timestamps, and aggregating intermediate CDRs or segmented SMS messages) before reconciliation.
• Advanced Analytics: Spark’s integration with MLlib allows for the application of machine learning techniques to detect anomalies or patterns in CDR data that could indicate issues such as revenue leakage or fraud.
• Automated Matching: Spark can be used to automate the matching process based on multiple criteria, such as caller, called, timestamp, and imsi, and flagging records that do not match for further analysis.

Conclusion

CDR reconciliation between MSC and IN is a vital process for telecom operators, enabling them to maintain accurate billing, detect fraud, and ensure compliance. Although challenges like the absence of a unique correlation ID and time differences exist, these can be mitigated with proper data processing and advanced tools like Apache Spark. By adopting such big data solutions, telecom operators can not only streamline the reconciliation process but also gain valuable insights to drive business decisions.

Join Our Upcoming Webinar: Preventing Revenue Leakage: Core vs. Intelligent Network Reconciliation. 

Want to learn more about data reconcialition between core and intelligent network? Join our insightful webinar

Key Topics:

• The importance of reconciling data between core network elements (MSC, SMS-C, GGSN, SGSN) for revenue assurance.
• Common challenges faced during Voice, SMS and Data reconciliation.
  
• Practical demonstrations on how to reconcile data effectively between core network elements and the Intelligent Network (IN) to pre

📅 Date: December 17th, 2024
⏰ Time: 10:30 AM UTC+1
🌐 Platform: Zoho

Register Now!