© 2022 Black Swan Telecom Journal | • | protecting and growing a robust communications business | • a service of |
Email a colleague |
February 2015
International call bypass is one of telecom’s toughest fraud problems.
The fraud is primarily accomplished through SIM boxes equipped with dozens to hundreds of SIM cards that disguise international calls toward wireless operators as local subscribers making domestic phone calls.
These SIM boxes are causing major revenue damage to wireless operators in many countries. Interestingly though, no one quite knows the actual volume and magnitude of the revenue lost from SIM boxes because detecting the bypass is an art, and not yet a science. Compounding the problem, the fraudsters constantly develop new techniques to avoid detection.
Operators are not the only losers by the way. Many countries impose a tax on telecom services, so the governments in those countries lose tax revenue when calls are illegally bypassed.
One of the innovators in the fight to curtail SIM box fraud is a U.S.-based company, LATRO Services. Their company’s CEO, Lex Wilkinson, now joins us to give a backgrounder on SIM box detection techniques and tells us about LATRO’s new technology, Protocol Signature™ detection data analytics based on network protocol analysis that detects SIM boxes as they sign onto the network.
Dan Baker: Lex, what percent of the countries in the world experience this SIM box fraud? |
Lex Wilkinson: It’s difficult to identify the exact percentage of countries, Dan, but I think that the percentage is very high. A major misconception is that SIM box fraud happens in countries with very high termination rates, but based on our experience, we’ve seen it occurring even in countries where the rates wouldn‘t be considered very high.
The money-making opportunity for the fraudsters is the differential between the international termination rate and the local termination rate. So even in countries where that is only a few cents, there’s still enough margin for fraudsters to make money.
The biggest hotbeds for SIM box fraud that we see are Africa, the Middle East, Central and South Asia, Latin America, the Caribbean, and Eastern Europe.
Gee, you didn‘t leave many continents out of the picture. I’m curious: how do these fraudsters get into the business anyway? |
For a fraudster to set up a profitable SIM box operation, the proper infrastructure needs to be there. For instance, there needs to be reasonably good Internet service such as DSL, satellite, or fixed wireless like WiMAX. We’re also finding that some of the wireless 3G and 4G data services deployed in the developing world are good enough to support termination of VoIP traffic to the SIM Boxes. There also needs to be a plentiful supply of SIM cards because the fraudsters need to access a large volume of SIM cards on a regular basis.
We usually see fraud operations set up in urban areas within developing countries. It’s rare — but not unheard of — to see SIM box operations in rural areas: broadband Internet connections and readily available supplies of SIM cards tend to be in the bigger cities. Often times, access to the SIM cards in volume is made possible through dealer fraud. A lot of countries require registration to get a SIM card, such that buyers need to show an ID in order to purchase a SIM. In our experience. However registration controls do little to stop SIM Boxes, but result in a criminal business around fake and stolen IDs used to acquire SIM cards.
Lex, I’m eager to learn about your new SIM box detection technology. |
Well, I think the most effective way of explaining our solution is to review the evolution of SIM box detection solutions. Let me walk you through the strengths and weaknesses of the two primary solutions used in the past 10+ years — test call generation (TCG) and fraud management systems (FMS). Then I’ll discuss some of the advantages of what our new technology — Protocol Signature™, data analytics based on network protocol analysis — brings to the table.
When the SIM box bypass problem was first identified in the late 2000’s, test call generation was the first detection technique that proved effective. The idea behind test calls is to set up test phone numbers in your network and make calls to those test number from lots of different countries, through many different interconnect voice routes around the world. In this way, you can find out where the grey routes are originating and the paths they use to reach SIM Boxes in your country.
Test Call Generation is all about probability. The more routes and the more test calls you make, the higher the chances of finding SIM boxes. Once you find routes that have a high volume of SIM box terminations, you can focus your call campaigns on those routes in order to maximize detections as much as possible.
Test Call Generation technology worked very successfully for many years. Yet in the last two to three years we feel the effectiveness of that approach has dropped off significantly. There are a few reasons for that. First, the SIM boxers have figured out how to avoid detection by test calls. For instance, they perform analysis on the voice call traffic coming toward their SIM boxes. Based on usage patterns and other patterns, they can determine which calls are real subscriber calls and which are originating from a test call generation system.
Then they can either block the test calls and prevent them from reaching the SIM box to begin with, or reroute the calls to a legitimate route in order to avoid detection.
We’ve also heard cases of fraudsters allocating pools of their SIM Box cards to be sacrificed. That is, they allow these allocated SIM cards to be detected in order to make the losing wireless operators feel like their controls are producing adequate results. This is really only a diversion. Meanwhile, other undetected SIM cards are driving the bypass revenue losses.
Bottom line: test calls are a technology and methodology that’s now well-understood by the fraudsters. You still need it. We don‘t recommend anyone stop using test calls. Test calls do get some results and allow a wireless operator to profile their interconnect partners and understand who’s sending bypass traffic toward their networks.
Another solution traditionally used to detect SIM boxes is the Fraud Management System (FMS), an enterprise wide data analysis platform that works well in detecting many different types of fraud. In SIM box detection, the FMS uses Call Data Records (CDRs) to create usage-based analysis profiles that detect SIMs being used in SIM boxes versus those used in legitimate subscriber handsets.
FMS and similar CDR Analysis platforms have been effective in detecting SIM boxes, but in recent years, fraudsters have figured out ways to evade usage profile detection. For instance, SIM Box manufacturers have developed something called HBS — Human Behavioral Simulation software — that allows the SIM boxer to simulate the behavior of real mobile subscriber behavior. HBS techniques involve automating features on the SIM Box such as SMS messaging, self-calling, and international dialing in order to frustrate detection algorithms used by FMS and CDR analysis.
Now in both cases — test calls and the FMS — by nature of their methodology, the fraud has been committed by the time you detect it. So you are already losing money before the detections occur. This is a major limitation.
At LATRO Services, we figured there must be a better way to attack this problem. Our expertise and background is in mobile network architectures and systems: we know a lot about protocol signaling and systems integration of network elements within a wireless network.
So we developed algorithms that use information available within the network itself to help detect SIM box fraud. In fact, we can detect SIM Boxes on the network at the moment they connect. So as soon as the SIM box is powered up and the SIMs are inserted, we can do the detection regardless of usage analysis and test calls.
As a wireless operator, you’re never quite sure when a SIM box is going to turn on within your network. You have to always be looking in real-time for the patterns in individual protocol messages of all the wireless devices on your network. So we developed a technology that allows us to flag what we call the Protocol Signatures of SIM box devices vs. the devices a normal mobile subscriber uses.
Now the information we are processing and analyzing is based on signaling data that is not available in CDRs. So we feel we are offering something unique to the market. Something that test calls, FMS, and CDR Analysis cannot leverage.
Lex, this rundown of the different kinds of detection platforms is very valuable. And I would love to hear the FMS and test call vendors chime, too, so I invite them to comment. Where do you feel operators need to modify their methods of combating the SIM Box? |
I think too many operators assume their current suite of tools is sufficient to detect and control the SIM box problem in their network. Yes, they are getting some results, but my hunch is the results are not really mitigating the problem effectively.
Looking at this from a revenue perspective, if a wireless operator is doing a good job of stopping SIM boxes on its network, the volume of international traffic to that network should be increasing. We know there should be a strong correlation here: if you do a good job at SIM box detection, yet your international call revenue is decreasing, there’s a disconnect somewhere.
In fact, many operators are frustrated by revenue declines and are looking for SIM Box control solutions that go beyond what’s available today. And that’s where we are looking to add value.
For an effective mitigation strategy, you need to approach the problem from multiple angles. For example, though SIM registration and distribution control is a good strategy, it needs to be complemented by other strategies. We would never say our solution alone is a silver bullet. But we do feel we have an innovative technology that strongly complements and even outperforms other solutions currently in use. Clients that implement our solution as part of their overall bypass fraud control strategy, gain incremental revenue to their top lines.
Now I understand that recently some fraudsters have been arrested thanks to results from your platform. |
If you scan Google under “SIM box arrests” you will find a few stories of SIM boxer arrests in countries like Senegal, Ghana, Haiti, and Morocco.
And yes, some of our clients have asked us to do investigation work. Using our technology, we are able to physically locate SIM Box equipment within the wireless network. Then in cooperation with our clients, we support local police to make arrests. We use our analytics platform to calculate initial location estimates based on network data and tower locations, then we have RF-based equipment used in drive testing to pinpoint the exact location of the equipment via direction-finding techniques.
In cases where the fraud has been prosecuted and equipment confiscated, we’ve seen fraud operations with tens of thousands of SIM cards found on-site. These are SIM cards that were blocked by control techniques like test calls and FMS. So all the fraudsters do is throw away the blocked SIM cards and replace them with new ones in their SIM Box equipment. Clearly, the fraudsters expect a large number of SIMs to be blocked and they know how to get their hands on more and more SIM cards.
The quicker you can detect and block, the higher the cost to the fraudster — in both money and potential jail time.
Lex, thanks for a timely briefing and advice in this important SIM box area. And good luck in your quest to find better ways to combat this fraud. |
Copyright 2015 Black Swan Telecom Journal