skip to main |
skip to sidebar
Overview of Telecom Mediation
Telecom Mediation is a system which acts as an interface or a middle layer between network elements and billing system/down stream applications like Fraud Management Systems, Revenue Assurance, Data Ware house etc.
The functionalities that are performed by Mediation are as follows.
· Mediation collects the CDRs from the network elements (MSC, SGSN, GGSN etc).
· Segregation of CDRs (separation of different kinds of CDRs).
· Mediation performs data enrichment and format conversion.
· Preprocessing and Post processing of CDRs.
· Applies business rules such as Filtration, Aggregation, Correlation, and Duplicate check etc on the CDRs.
· Sending SNMP traps in case of errors during the parsing, conversion, distribution and processing of CDRs in the mediation system.
· Distribution of BDRs/CDRs to the billing systems or down-stream applications
CDR Collection
Mediation collects the usage records/CDRs from the network elements. Mediation supports the collection of the usage records using various protocols that are widely supported namely FTP, SFTP, FTAM, X.25 Serial interface, MTP, SNMP. In the GSM networks the widely used interfaces to collect the CDRs are FTP/SFTP and FTAM. SFTP provides a secured way of collecting the CDRs as it uses SSL (secured socket layer) which encrypts the data while transferring the data in a secured way.
Preprocessing, Parsing and Validation
Preprocessing is a process of removal of headers/trailer information for further processing. Preprocessing also involves splitting the fields of the CDR, adding additional fields in the CDR required for processing. This may involve file or data base look ups to populate the route ID’s etc.
The collected or preprocessed CDR is parsed to identify whether all the required fields are present in the CDR and validated to the switch format to check whether it meets the specification. This involves checking the length of the CDR, usage field type (type of CDR) and fields are in proper order.
NE to down stream application mapping
In order to apply any kind of business rules (processing functions) a mapping has to be created between the NE and the down stream applications. These business rules are applied on the stream either on the Input side (NE format) as IPF (input processing function) or on the Output side (Application format) as OPF (output processing function). Now once the mapping is made the mediation system understands which NE CDRs are to be routed/ distributed to which down stream applications once the processing process is completed.
Business Rules (Processing Functions)
There are different types of business rules that are used in mediation.
· Sequence Validation
· Filtration
· Aggregation
· Duplicate CDR Check
· Correlation
· Time Gap Detection
· Trunk Analysis
Sequence Validation
Sequence validation is applied on the raw CDRs received from the Network Element. This business rule is used to find out whether all the CDRs received from the NE are in order. This helps in uncovering any frauds that might have happened due to suppressing of CDRs after they are generated in the NE. It also helps in knowing whether all the CDRs are received and there is no CDR loss.
The above figure shows the CDRs that are generated by a MSC. There are four CDRs in the CDR file. The CDR with sequence number 0003 is missing. The mediation records the missing sequence in the error logs and generates the SNMP trap for the same saying that the CDR is out of sequence and the expected CDR sequence was 0003 but received CDR sequence number is 0004.
Filtration
Filtration is a business rule used for eliminating the CDRs which are not required for further processing. There are certain kinds of CDRs which have to be filtered from the mediation system and should not be sent for billing. These CDRs may be of the following type.
· Zero duration Call CDRs
· SMS terminated / Mobile terminated CDRs (as incoming SMS/call is free)
· Toll free numbers
Aggregation
Aggregation is a business rule used for stitching the call legs of the long duration calls on the basis of partial CDRs to convert it into a single CDR. Here long duration calls doesn’t mean STD/ISD calls. The network element (MSC) is configured to generate a CDR for configured time duration say 10 minutes. If the call goes on for more than 10 minutes two call legs (CDRs) are generated for the same call session. These CDRs have to be stitched in order to make them as a single CDR. These individual CDRs are called as partial CDRs and the combined CDR is called as the aggregated CDR. This stitching is necessary because it has to be rated and billed as a single event rather than multiple events as it was a single call made by the user.
Let us take an example of a GSM network where in a user makes a call which is of say 50 minutes. The configured duration in the MSC (Network Element) is say 10 minutes. Now there are 5 partial CDRs generated. They have to be stitched together to form a single CDR. The above figure depicts the partial CDRs and their corresponding durations. The mediation once starts parsing the CDR file searches for the first partial CDR (first call leg) and starts stitching the CDRs until it gets the final or complete partial CDR. All the partial CDRs carry some value to denote whether they are first, intermediate or final partial CDR. Once the final partial CDR is found, mediation system constructs the aggregated CDR with all the necessary fields with the summation of the duration field. The aggregated CDR looks as shown below
Duplicate CDR Check
Duplicate CDR Check processing function is applied in order to find out whether there are any duplicate CDRs or duplicate CDR files. Duplicate CDR checks help in elimination of the CDRs or the CDR files which are duplicate and should not be processed as they are already processed. The combinations of some of the CDR fields are used to identify whether the CDR is duplicate or not.
o Unique Transaction ID
o Mobile Originating Number
o Mobile Terminating Number
o Date/Time…
The CDR file sequence number is maintained in order to check whether the CDR file itself is duplicate or not.
The processed CDRs are kept into the CDR data store. Whenever a CDR is received this data store is compared to find out whether the CDR is duplicate or not. The CDR store holds the CDR records for a configurable period of time.
Correlation
Correlation is a process of comparing data from two or more network elements.
Correlation is a business rule in mediation that is generally applied on the CDRs received from one or more Network elements. These CDRs have to be stitched together in order to get the billable information from them. Correlation means comparing some of the fields which are common across the network elements and extract the information (other fields) needed and stitch them to form a single field.
Let us take an example of a GPRS network to demonstrate the correlation feature.
Whenever we use our GSM mobile phone to access the GPRS services like browsing a website or downloading some information from a portal, the radio related information is generated in SGSN and the IP related information is generated in the GGSN. The SGSN generates the radio related information including the Calling number field which is called as the S-CDR and the GGSN generates all the fields related the QOS (Quality of service) like download rate, bandwidth and the temporary IP used etc. which is called as G-CDR. These S-CDR and G-CDR will have a common field like transaction id to identify that they have to be correlated and stitched to form a single CDR as both have some information which are needed to make the call billable.
Time Gap Detection
Time Gap detection business rule is used in mediation is a process of finding out the time gap between two consecutive usage records or CDRs. This helps in finding out whether any of the usage events are missing. Generally a configured amt of time exists between the two consecutive CDR generations. The time between the two consecutive CDRs is calculated and compared with the standard/configured time duration. If the time duration between the two consecutive CDRs generated is greater than the configured time duration, an alarm or alert is generated to notify it which may be a cause of missing CDRs or suppressed CDRs.
Trunk Analysis
Trunk Analysis is generally done in order to check whether the incoming and the outgoing trunk fields in CDR match to that of the incoming/outgoing trunk in the telecom network. This process is done through either a table lookup or file lookup. All the incoming trunks and the outgoing trunks configured in a telecom network are listed in the tables. They are compared and if there are any different trunk values in the CDR other than the ones in the table a report is generated for the violation and later checked by Fraud management systems for any frauds that may have happened.
Data Enrichment and format conversion
Mediation main functionality is to enrich the data and format conversion. Mediation supports ASN.1, binary and ASCII data. Any CDR received in any form (ASN/Binary and ASCII) can be converted to other formats.
Format conversion from ASN.1 to ASCII
Format conversion from ASN.1 to Binary
Format conversion from Binary to ASCII
Format conversion from Binary to ASN.1
Similarly other combinations are also possible. The input CDR (NE) and the output CDR fields (Billing/Down-stream applications) will be different. Say there are 20 incoming fields in input and there might be only 10 fields needed for billing.
Post processing and Distribution
Post processing involves adding trailers and header information in the CDR/BDR, adding additional fields that may be required by billing etc. The distribution process in mediation is a process of sending the CDRs/ BDRs to the billing or down-stream applications over a interface which use protocols like SFTP, FTP, FTAM etc.
Telecom Knowledge
Overview of Telecom Mediation
Overview of Telecom Mediation
Mediation
Telecom Mediation is a system which acts as an interface or a middle layer between network elements and billing system/down stream applications like Fraud Management Systems, Revenue Assurance, Data Ware house etc.
The functionalities that are performed by Mediation are as follows.
· Mediation collects the CDRs from the network elements (MSC, SGSN, GGSN etc).
· Segregation of CDRs (separation of different kinds of CDRs).
· Mediation performs data enrichment and format conversion.
· Preprocessing and Post processing of CDRs.
· Applies business rules such as Filtration, Aggregation, Correlation, and Duplicate check etc on the CDRs.
· Sending SNMP traps in case of errors during the parsing, conversion, distribution and processing of CDRs in the mediation system.
· Distribution of BDRs/CDRs to the billing systems or down-stream applications
CDR Collection
Mediation collects the usage records/CDRs from the network elements. Mediation supports the collection of the usage records using various protocols that are widely supported namely FTP, SFTP, FTAM, X.25 Serial interface, MTP, SNMP. In the GSM networks the widely used interfaces to collect the CDRs are FTP/SFTP and FTAM. SFTP provides a secured way of collecting the CDRs as it uses SSL (secured socket layer) which encrypts the data while transferring the data in a secured way.
Preprocessing, Parsing and Validation
Preprocessing is a process of removal of headers/trailer information for further processing. Preprocessing also involves splitting the fields of the CDR, adding additional fields in the CDR required for processing. This may involve file or data base look ups to populate the route ID’s etc.
The collected or preprocessed CDR is parsed to identify whether all the required fields are present in the CDR and validated to the switch format to check whether it meets the specification. This involves checking the length of the CDR, usage field type (type of CDR) and fields are in proper order.
NE to down stream application mapping
In order to apply any kind of business rules (processing functions) a mapping has to be created between the NE and the down stream applications. These business rules are applied on the stream either on the Input side (NE format) as IPF (input processing function) or on the Output side (Application format) as OPF (output processing function). Now once the mapping is made the mediation system understands which NE CDRs are to be routed/ distributed to which down stream applications once the processing process is completed.
Business Rules (Processing Functions)
There are different types of business rules that are used in mediation.
· Sequence Validation
· Filtration
· Aggregation
· Duplicate CDR Check
· Correlation
· Time Gap Detection
· Trunk Analysis
Sequence Validation
Sequence validation is applied on the raw CDRs received from the Network Element. This business rule is used to find out whether all the CDRs received from the NE are in order. This helps in uncovering any frauds that might have happened due to suppressing of CDRs after they are generated in the NE. It also helps in knowing whether all the CDRs are received and there is no CDR loss.
The above figure shows the CDRs that are generated by a MSC. There are four CDRs in the CDR file. The CDR with sequence number 0003 is missing. The mediation records the missing sequence in the error logs and generates the SNMP trap for the same saying that the CDR is out of sequence and the expected CDR sequence was 0003 but received CDR sequence number is 0004.
Filtration
Filtration is a business rule used for eliminating the CDRs which are not required for further processing. There are certain kinds of CDRs which have to be filtered from the mediation system and should not be sent for billing. These CDRs may be of the following type.
· Zero duration Call CDRs
· SMS terminated / Mobile terminated CDRs (as incoming SMS/call is free)
· Toll free numbers
Aggregation
Aggregation is a business rule used for stitching the call legs of the long duration calls on the basis of partial CDRs to convert it into a single CDR. Here long duration calls doesn’t mean STD/ISD calls. The network element (MSC) is configured to generate a CDR for configured time duration say 10 minutes. If the call goes on for more than 10 minutes two call legs (CDRs) are generated for the same call session. These CDRs have to be stitched in order to make them as a single CDR. These individual CDRs are called as partial CDRs and the combined CDR is called as the aggregated CDR. This stitching is necessary because it has to be rated and billed as a single event rather than multiple events as it was a single call made by the user.
Let us take an example of a GSM network where in a user makes a call which is of say 50 minutes. The configured duration in the MSC (Network Element) is say 10 minutes. Now there are 5 partial CDRs generated. They have to be stitched together to form a single CDR. The above figure depicts the partial CDRs and their corresponding durations. The mediation once starts parsing the CDR file searches for the first partial CDR (first call leg) and starts stitching the CDRs until it gets the final or complete partial CDR. All the partial CDRs carry some value to denote whether they are first, intermediate or final partial CDR. Once the final partial CDR is found, mediation system constructs the aggregated CDR with all the necessary fields with the summation of the duration field. The aggregated CDR looks as shown below
Duplicate CDR Check
Duplicate CDR Check processing function is applied in order to find out whether there are any duplicate CDRs or duplicate CDR files. Duplicate CDR checks help in elimination of the CDRs or the CDR files which are duplicate and should not be processed as they are already processed. The combinations of some of the CDR fields are used to identify whether the CDR is duplicate or not.
o Unique Transaction ID
o Mobile Originating Number
o Mobile Terminating Number
o Date/Time…
The CDR file sequence number is maintained in order to check whether the CDR file itself is duplicate or not.
The processed CDRs are kept into the CDR data store. Whenever a CDR is received this data store is compared to find out whether the CDR is duplicate or not. The CDR store holds the CDR records for a configurable period of time.
Correlation
Correlation is a process of comparing data from two or more network elements.
Correlation is a business rule in mediation that is generally applied on the CDRs received from one or more Network elements. These CDRs have to be stitched together in order to get the billable information from them. Correlation means comparing some of the fields which are common across the network elements and extract the information (other fields) needed and stitch them to form a single field.
Let us take an example of a GPRS network to demonstrate the correlation feature.
Whenever we use our GSM mobile phone to access the GPRS services like browsing a website or downloading some information from a portal, the radio related information is generated in SGSN and the IP related information is generated in the GGSN. The SGSN generates the radio related information including the Calling number field which is called as the S-CDR and the GGSN generates all the fields related the QOS (Quality of service) like download rate, bandwidth and the temporary IP used etc. which is called as G-CDR. These S-CDR and G-CDR will have a common field like transaction id to identify that they have to be correlated and stitched to form a single CDR as both have some information which are needed to make the call billable.
Time Gap Detection
Time Gap detection business rule is used in mediation is a process of finding out the time gap between two consecutive usage records or CDRs. This helps in finding out whether any of the usage events are missing. Generally a configured amt of time exists between the two consecutive CDR generations. The time between the two consecutive CDRs is calculated and compared with the standard/configured time duration. If the time duration between the two consecutive CDRs generated is greater than the configured time duration, an alarm or alert is generated to notify it which may be a cause of missing CDRs or suppressed CDRs.
Trunk Analysis
Trunk Analysis is generally done in order to check whether the incoming and the outgoing trunk fields in CDR match to that of the incoming/outgoing trunk in the telecom network. This process is done through either a table lookup or file lookup. All the incoming trunks and the outgoing trunks configured in a telecom network are listed in the tables. They are compared and if there are any different trunk values in the CDR other than the ones in the table a report is generated for the violation and later checked by Fraud management systems for any frauds that may have happened.
Data Enrichment and format conversion
Mediation main functionality is to enrich the data and format conversion. Mediation supports ASN.1, binary and ASCII data. Any CDR received in any form (ASN/Binary and ASCII) can be converted to other formats.
Format conversion from ASN.1 to ASCII
Format conversion from ASN.1 to Binary
Format conversion from Binary to ASCII
Format conversion from Binary to ASN.1
Similarly other combinations are also possible. The input CDR (NE) and the output CDR fields (Billing/Down-stream applications) will be different. Say there are 20 incoming fields in input and there might be only 10 fields needed for billing.
Post processing and Distribution
Post processing involves adding trailers and header information in the CDR/BDR, adding additional fields that may be required by billing etc. The distribution process in mediation is a process of sending the CDRs/ BDRs to the billing or down-stream applications over a interface which use protocols like SFTP, FTP, FTAM etc.