SS7 Networks:Rich Data Stream Holds Promise for Revenue Assurance
Michelle L. Hankins
02/01/2001
Carriers are finding ways to use information gleaned from the signaling system 7 (SS7) network to prevent revenue leakage, better manage their networks and combat fraud, but new techniques have to be developed before the data can be used to its full potential.
SS7 networks carry the information carriers need to set up and tear down calls; service providers can capture that data and build a thorough and complete record about every call attempt on the network. SS7 can also help a provider determine all the necessary information for rating—calling party number, the party called, duration, and the time the call started and stopped. It also recognizes the type of call—including fax, voice mail (or checking voice mail), point-of-sale validations and credit card transactions. It even can identify ISP calls.
“I think this is an emerging area,” says Paul Florack, director of network services at Illuminet, an SS7 network provider. “I don’t think there is a widespread use of this information yet today, but it is definitely growing in interest because of the breadth of information you can pull out of the SS7 network.”
SS7 for Revenue Assurance
To use SS7 data for revenue assurance, a provider collects the SS7 call records and compares them against automatic message accounting (AMA) records to verify the network’s performance. If the data sets are out of sync during the audit, the information can be used to pinpoint weaknesses in the network. “It’s a parallel data collection exercise,” says Rod McIntosh Shand, vice president of research at Botting Systems Inc.
Typical mistakes usually include calls that don’t generate an AMA record, which leads to lost billing opportunities. This happens when a numbering plan area (NPA) is incorrectly marked as “local” in a table. In this case a provider might find SS7 records for calls going to that NPA with no matching AMA records going to that NPA.
“Having made sure that the records coming out of the switch are correct,” Shand says, a provider can check the various phases of its billing process against the SS7 output by writing its own AMA records. To the billing system, these AMA records can look like test calls or real calls. Once a provider has generated these records, it can put them through the billing system and compare the actual bill against the expected results to detect any errors in the billing system. After a provider does this once, it can automate the AMA record generation to test the system again if any changes are made that might impact the process. But writing AMA records is a truly tedious process, which makes this kind of testing problematic.
PricewaterhouseCoopers (PwC) Partner Luke Sayers, who leads the Revenue Maximizer Services practice, believes that though SS7 is a valuable source of information for revenue assurance, the industry hasn’t learned to mine it deeply enough.
“The strategy is definitely to move further up the food chain from test call generators to some of this SS7 technology that can look across both networks and give you real-time information,” Sayers says.
Because SS7 data originally was designed to measure network performance (see “The Evolution of SS7 Data,”), Sayers believes it is still just too difficult for business and revenue assurance teams to translate the high volume of “engineering speak” that rolls off the SS7 network. “It’s great to have all this fancy network output, but they have not done a very good job at bridging the gap between the output and what the business needs to be able to make sense of it,” he says. “People are not getting the biggest bang for their buck because … it’s too difficult to really be able to analyze the information.”
PwC worked with one SS7 tool vendor to measure SS7 data for revenue assurance. Trying to compare access and egress information off the SS7 network and combine it to make a call record was extremely difficult, says Jonathan Bernstein, manager of Revenue Maximizer Services at PwC. “It was not an easy, one-to-one relationship, and it took a great deal of time to associate the access piece of the call to the egress piece, so we couldn’t get a full picture of a call record.”
Matching the records also is a difficult process. “Because the switch clock will drift several seconds and the SS7 links have their own independent links which will drift as well, it was next to impossible to compare one call record from the SS7 link to one call record from the switch. We wanted to do it as a mass edit—sort of as a large reconciliation—but even being able to reconcile one call record to one call record was very difficult,” Bernstein says. “We didn’t get a warm and fuzzy feeling that the SS7 was the accurate source, because we were having so many discrepancies between that and reconciling it to the switch.”
PwC, however, was more successful using SS7 data regarding interconnection issues. “It’s a great way to confirm what the other carriers are billing for,” Bernstein says.
Intercarrier Billing
Many providers already know the importance of SS7 for intercarrier bill verification. “It’s a low-cost way of getting an intercarrier billing system,” says Kevin Keough, vice president of corporate development at Inet Technologies Inc. “Using SS7 you can get as much, if not more, detail than you can using switch and call detail records [CDRs]. “SS7 is an international standard, so you don’t have to customize your billing system to accommodate different switch vendors’ CDR formats.”
Keough says that even if carriers, particularly ILECs, have a carrier-to-carrier billing system based on CDRs, they may still question that system’s integrity. “At a minimum, our system provides a parallel alternative feed by which you can compare what you’re getting out of the SS7 network in terms of billable records, and what you’re getting out of the switch CDR-based system for billable records. It will very quickly determine if there is a revenue assurance leakage problem from the existing billing system.”
The newer competitive carriers, Keough says, have avoided spending money on carrier-to-carrier billing systems. “If they are going to buy a billing system, they are going to spend it on an end-user billing system. They don’t want to go spend millions of dollars to buy a carrier-to-carrier billing system just to verify the invoices they are receiving from the incumbent carriers for terminating their traffic. They are looking for a low-cost way to verify the bills they are receiving from the incumbent carriers.”
One SS7 tool can write summary reports detailing call origination, termination and duration. “So when you get the bill from the incumbent carrier for half a million dollars, you can look at it and have data to either support or contest that bill,” Keough says.
Agilent’s Alan Thomson, product manager of the company’s acceSS7 billing product, says, “We have some customers who have done studies using some of our business intelligence applications who reckon they have seen 20–30 percent discrepancies in their bills. We’ve actually done trials with our product with one of our customers where we put our product versus five switch types over a month period and we were as accurate, if not better than, those switch records.”
Switches don’t always align, either, Thomson says. “All it takes is wrong configuration in a switch, wrong configuration in OSS, and you start running into discrepancies.”
“Most carriers will still generate billing information off the switch, but they don’t have a real good mechanism to verify that their switch is generating the right amount of information,” Illuminet’s Florack says, so bill verification on both incoming and outgoing bills is critical.
“It’s basically a neutral source of information, especially from an intercarrier billing perspective. Rather than a carrier’s switches generating the data, you pull a message stream that both networks see to create a CDR that can provide a good alternative to the switched CDR,” Florack says.
Intercarrier Billing Issues
A typical problem with intercarrier billing might occur when the switch for an intercarrier trunk group doesn’t generate records. SS7 records can identify that event for that specific trunk group, alerting the provider that there’s a switch problem at that point on the network. The provider can then determine where to look for the problem in the switch’s tables, says Shand at Botting Systems.
He recalls one case in which a LEC made some changes to its switch for equal access and lost about a month’s worth of long-distance call records. “The changes they were making—as far as they knew—should not have affected toll recording at all, but they did. It was a side effect. So you have a general problem that if you make changes to a switch’s configuration—and particularly anything that touches on AMA recording—you really ought to test it very carefully before you allow it to run unsupervised.”
SS7 data can also verify contract compliance. “We discovered in one case that whereas the contract said that the bill per call would be calculated on the talk time—that is, from the time the call was answered until the time somebody hung up—it was actually being billed on circuit occupancy time—which is from the time the call was started, until the time the call was ended—whether anybody answered or not. And even if they did answer, it included the ringing time and the setup time,” Shand says. “Over a month, that can add up. All it means is whoever configured the billing system did not configure it according to the contract that was signed, so what you’re looking at to an extent is contract compliance.”
Network Performance
Network performance is critical because poor quality of service (QoS) can also lead to revenue leakage. A wireless switch once took 10 seconds to decide what to do with a call, Shand says, which meant the caller had to put up with 10 seconds of silence.
“They could be losing quite a lot of revenue from people who are abandoning calls because one of their interconnect carriers, or indeed, one of their own switches is taking an awfully long time to decide what to do with the call,” he says.
Other QoS issues that might turn users off include fast busy signals or busy circuit messages that occur as a switch searches for an open circuit. “It’s very probable that the switch will not generate a record for this—why would it? However, there will be an SS7 record for it, and it will show that calls are dropped. So what you can do is look at this and find out which of their interconnect partners are actually carrying the traffic that they say they can carry for you, and which ones have underprovisioned their networks so that they can’t actually offer the service that you’ve contracted them to provide,” Shand says.
It’s important for carriers to determine why a call failed, especially if it’s a problem on another carrier’s network. Because customers don’t understand the call’s transmission path, they tend to blame the carrier to which they subscribe, Florack says. Using SS7 to determine where the call dropped helps the subscriber’s carrier prevent customer churn.
A company can also use SS7 data for optimum network configuration, a major issue where data traffic travels across the PSTN.
Managing Data Traffic
The rise in Internet traffic has changed the dynamics of the PSTN. Providers must now be able to manage network capacity to keep up with increasing demand—and longer call duration, especially when a user is surfing the Web.
“If you know what kind of traffic is data traffic, then the best thing you can do with it is to get it off the voice network as fast as you can, as close to its point of origin as you can, and handle it as data,” Shand says.
Botting Systems performed a network analysis for an ILEC that was experiencing a rapid increase in Internet traffic. Botting pulled SS7 call records to pinpoint where and when the ISP Internet traffic was originating and terminating. As a result, the ILEC redesigned part of its network to optimize Internet traffic to ISPs. The carrier estimated that it saved about $28 million, Shand says, putting its benefit-to-cost ratio at about 20-to-1.
Keough at Inet Technologies explains how providers divert data traffic off the circuit-switched network onto an IP packet or an ATM cell network: When a user connects to the Internet, a modem calls a remote access server (RAS) to negotiate the connection. The call goes through the switch closest to the user, and on to other switches, which ties up circuits, before the call eventually reaches the RAS or the modem pool. Surfing the Internet takes an enormous amount of circuit capacity that can otherwise be used for voice calls.
“Instead of that RAS being the last point that your call touches before it hits the Internet, we’re trying to move that point as close to your house as possible,” Keough says. “So in effect your call, instead of going through these four switches, goes through the first switch and automatically gets piped out of that first switch into the RAS and directly to the Internet, thereby relieving all those other switches along our chain to carry voice calls.”
“What a lot of carriers were doing was throwing a lot of money at access tandems and trunking backbone,” Florack says, “rather than taking a look at what’s causing this [congestion] to more efficiently route where the traffic comes in and out. At the moment, you may call from a local exchange through an access tandem to a CLEC, whereas if that CLEC is doing a lot of ISP traffic, it may be that you want to put direct trunks between the CLEC and your local exchange as opposed to using up your tandem’s capacity.”
SS7-Based Billing
Products that allow providers to measure SS7 trunk usage are popular in North America, Agilent’s Thomson says. In addition to charging based on the duration of a call, a provider might bill for use of the SS7 network for terminating that call.
This is particularly important in the wireless arena, Thomson says, where a traveling subscriber might turn on his phone while in another part of the country. While a call may never be placed to initiate use of a voice trunk, the SS7 network would still be used to determine roaming rights. A provider might want to recover cost of use of the SS7 network in such instances.
Short message services (SMS) are beginning to tax the SS7 network. When one mobile phone user sends text messages to another wireless subscriber, a transaction request is sent across the SS7 backbone to the terminating point. Those messages don’t have anything to do with voice, but they use up a carrier’s network bandwidth before it crosses to another carrier’s network.
SMS comprise up to 30 percent of international carriers’ SS7 traffic for which they have no accurate way to bill, Thomson says. They recover some of the cost through a fixed monthly charge, but don’t have enough information to determine whether their pricing schemes are accurate.
Fraud
Data from SS7 networks can also help carriers detect a fraudulent call while the perpetrator’s still on the phone, thus cutting revenue loss, Keough says. An example: A person steals a calling card and uses it to make calls to another country for 14 consecutive hours. A carrier may not realize the fraudulent act until it goes to collect its money. This is expensive because not only does the carrier not collect for those calls, it also has to pay the cost of network use, especially if it traverses other carriers’ networks.
While a switch-based fraud system generates CDRs after a call ends, SS7 CDRs can be generated at any point during the call. That information can include when the call is initiated, whether the call is terminating in another country, and whether the call exceeds a certain duration. If the carrier catches what it believes to be a fraudulent call, it cuts its losses by terminating the call. “The real-time aspect of SS7 is a dramatic benefit when we contrast it to systems that have historically been tied to latent-switch CDRs,” Keough says.
SS7 Versus Other Methods
Many suggest using SS7 data instead of generating test calls or other simulation to test networks. Why not simply consider regular subscriber calls as test calls? “Suppose you want to test your switching and your tariffs to South Africa,” Shand says. “Instead of making a test call to South Africa, for which you’re going to pay, you would just look at what happens to all the calls to South Africa by your subscribers—for which they are going to pay.” The AMA records could then be compared to the SS7 information for verification of the network’s performance.
But proponents of simulation or running test calls argue that SS7 networks can’t measure the quality of the voice trunk.
“It’s not intended to be an insertion and detection test tool,” Keough, says. “It’s meant to be an overall network monitoring system that can resolve things down to the circuit level, but it’s not meant to be [used for] inserting test tones into the trunk and seeing what they are on the other side.”
Keough says that test call tools that do that are basically point systems. “We’re the radar that sweeps across the whole city, and these guys can go figure out how much rain fell in your back yard.”
Extracting the Data
Whether SS7 data is used for more accurate intercarrier billing, monitoring network performance, managing data traffic or detecting fraud, carriers are calling on developers to create systems that sort through and use every detail provided by SS7 networks. “Billions of [SS7] messages fly across North America every day, and every one of them can tell you about something on the network,” Shand says.
“People know [SS7 data] exists but they don’t really know the enormous value,” he says. “The only way you would find out is by going into very careful detail in your collection analysis of that data. You really have to be adept at bit-twiddling in order to extract every ounce of data out of that data stream, but it is all there.”
In the past, SS7 monitoring systems simply managed network operations. They are now being deployed to pull data off the network for careful analysis for business and revenue purposes. But tools in this space still must evolve dramatically to let carriers make business sense out of network data.
Email this article
Add a comment
Printer version
Order reprints
RSS Feed
Bookmark article