Anti-Steering of Roaming: A Cause of Missed Revenue Targets - by Robin Burton ,Evolved Intelligence

Steering of roaming, as discussed in another briefing note, is a valuable tool that enables you to manage your relationships with your roaming partners effectively and dynamically. In today’s market it is a vital tool which helps you to control costs and manage quality for your roaming partners.

Naturally, those roaming partners whom you do not select, because of high cost or low quality, find that this activity represents a loss of revenue for them. Consequently they are tempted to interfere with the steering process. This interference can be called “anti-steering of roaming” or "roaming retention". It is often first suspected when steering targets are not met by the steering platform.

Anti-steering is proscribed by the GSMA. However there is no doubt that it is practised. In fact some vendors of steering platforms also own patents on various methods of anti-steering. This leads to the suspicion that these platforms are commercially available. We also recieve many requests for anti steering platforms, often disguised with descriptions such as "roam retention", "roamer attraction service", "inbound steering" or other euphemisms.

We usually suggest other ways to boost inbound roaming which do not lead to conflict with roaming partners and which generate genuine value for roamers.

Gateway Location Registers
Gateway Location Registers (GLRs) have become associated with anti-steering of roaming.

The “official” function of GLRs is to aggregate individual MSCs to reduce HPMN-VPMN signalling. As such a GLR function (if conformant to 3GPP standards) does not perform any steering of roaming or anti steering of roaming processes. However the architectural position of a GLR platform (in other words the hardware hosting GLR function) within a network is the ideal place to additionally host an anti-steering of roaming function.

It has been noted that some vendors have bundled anti steering of roaming functionality into their product offering, often in order to make a business case work. So although the ownership of a GLR does not necessarily mean that an operator is performing anti steering against you, it does mean that he could do so if he chose.

Some GLR vendors claim that their platforms do not perform "illegal" anti steering because they use "cached" signals from the roamers home network. This means that they save an "accept" signal from the home network and re-issue this signal to the handset the next time that it sends an update location request rather than allowing the handset's request back to the home network. Clearly this directly interferes with the home network's ability to direct the subscriber to chosen networks. In my view, therefore, it is clearly acting as an anti steering system even though the suppliers of such systems argue that they are not falsifying messages. The very fact that the vendors of such systems promote as a means of building inbound roamer share reveals that they also see it as an effective method of frustrating the home network's ability to direct their roamers.

Anti-Steering Methods
There are a number of ways in which anti steering can function.

Circumvention retention is a method where the anti-steering system deliberately emulates the behaviour of a subscriber trying to manually select a network. This is characterised by repeated attempts to select a network issued over a short period of time. Ultimately the steering platform, if operating to the GSMA guidelines, will recognise this as a genuine registration request. Kidnap Retention is a method whereby the anti-steering platform continually generates location update requests. This is done to retain the mobile’s registration data. The two methods are often used in combination.

Of course, from the steering platforms point of view, the behaviour may look innocent. The anti-steering platforms also try to disguise their presence as much as they can. Purely from the statistics generated by the steering platform, any individual event could be entirely innocent and be put down to coverage issues or user preference. However large numbers of events can lead to suspicion.

Anti steering platforms attempt to disrupt over the air “OTA” steering platforms by intercepting the binary SMS messages sent by the steering platform. The anti-steering platform may attempt to disguise this activity by sending back a spoof message receipt.

The GSMA PRD 89 formally defines anti steering as follows:
a) Any MAP_Update_Location or MAP_Update_gprs_Location messages generated by a VPMN, for which there is no simultaneous, linked and matching Air-Interface stimulus by the mobile device, is an Anti-SoR activity, or
b) Provided that the mobile device has maintained radio connection, the failure of the VPMN to generate an air interface message to the mobile device for each and every MAP Update Location response message, or TCAP/MAP error message received from the HPMN is an Anti-SoR activity. Furthermore, any deviation from conformance to 3GPP TS 29.010, when transferring the contents of the TCAP/MAP error message to the air interface message is an Anti-SoR activity, or
c) Any interference with HPMN "Over the air" (OTA) updating of SIM fields, for example by generation, discarding, manipulation or false acknowledgement, is an Anti-SoR activity.

However the immediately visible symptoms of anti-steering can often have innocent explanations

Detection
Usually suspicion begins when the steering platform fails to meet targets. A first step is to see whether the deviation can be due to other causes.

Network based steering is a statistical process. There are limits to the percentage targets that are achievable, even given a uniform distribution of network coverage from you various roaming partners. Add the variability of actual network coverage, and the steering percentage achievable is reduced. Please contact if you would like to know more about this and I will send you a graph showing realistically achievable steering targets.

If it still looks suspicious, the next step is to carry out a statistical analysis of the relevant SS7 signalling. In an Evolved Intelligence test, this includes tests on timing and velocity. The result of this usually delivers a very high level of confidence result.

However, for unequivocal proof, a test mobile needs to be used in the suspected territory. In the Evolved Intelligence test, we use an analysis of signalling from a friendly operator using our steering platform. This is compared with captured signalling from the test mobile together with a record of the availability of radio connection from various roaming partners. By looking at “both sides of the conversation” in this way we can generate an absolute proof of the use of anti-steering.

Is It A Good Idea To Use Anti Steering?
It can be very tempting to use anti steering. This offers the chance to get more traffic which you have not had to “pay for” by offering a lower price or by sending the partner operator more of your own roaming subscribers. It can be particularly tempting for new operators that do not have many outbound subscribers of their own which they can offer in “exchange” for inbound traffic.

We are often asked if we can offer “inbound steering” or “roamer retention”; all are essentially euphemisms for anti-steering. Some vendors of steering platforms have taken out patents on anti-steering techniques. In our roaming health checks we often see the tell-tale signs of it being used. Many operators suspect that they are the victims of anti-steering themselves (although often the actual poor performance that they see is because of the algorithm used by the steering platform itself). There is therefore little doubt that anti steering is being used. However is it a good idea to use it yourself?

Anti-steering will definitely bring you more traffic and more revenue. It is also likely that the value of this will outweigh the cost of running the anti-steering system. So, at least superficially, there would seem to be a business case for using it. The extra traffic and revenue will have been won at the expense of other operators and by deliberately frustrating the desires of your roaming partners.

You may or may not be worried by the ethical questions that this throws up. However you should certainly be worried about the potential reactions from your roaming partners. If your partners knew that you were using anti steering, they would certainly not be happy about it. They may react by adjusting their steering percentage targets to try and compensate for your anti-steering system. They may implement new anti-steering resistant platforms. Ultimately they may decide to cancel their roaming agreements with you altogether. The GSMA is also making increasingly strong comments about the practise. It is possible that operators could be ejected from the GSMA if “found guilty”.

Perhaps you still feel that the need for more roaming revenue is overwhelming. If so, before you implement anti steering, you should consider alternative ways of boosting revenue. We have seen how a full analysis of roaming traffic, for example through one of our roaming health checks, can show several ways in which revenue can be boosted. These include fixing missing roaming links and implementing dialled number correction. Outbound roaming can also be boosted through welcome messages and dialled number correction, making you a more attractive partner. All are highly effective ways of boosting revenue without running the risks inherent in anti-steering.


If you would like to know more, please get in touch: info (at) telecomadvisors.org