Sponsored networks and outbound roamer steering

There are different solutions allowing small operators to be able to inter-operate with quite large number of partner operators without the need of individual interworking agreements. Such solutions enable the new operators a global coverage with the support of the sponsor network.

Both sponsored and sponsor networks employ some products (which some might be called as roaming replicator), to make the message originated from sponsored network look like it is sponsor network’s traffic. Sponsor network route the traffic for the sponsored network properly too, which allows the sponsored network to access all the partners of the sponsor network.

As shown in the figure, when a subscriber attempts to camp to a network which receives service from a sponsor network, the registration request arrives from sponsor network to the HPLMN HLR. As the replicator products change both SCCP calling and VLR addresses with IMSI, from the point of HPLMN the subscriber is like being abroad in sponsor network country.

Issues with steering
Operators may prefer their outbound roamers to register to certain operators even though they may have roaming agreements with many operators in the same country. The motivation of having “preferred” partners can be cost (having discounts) or technical reasons (like better Camel / data support).

Steering solutions determine the country of the outbound roamer and try to steer the subscriber to a preferred partner using different methods. While one of the methods is network based steering which intercepts the subscribers’ registration messages, another method is SIM based steering which updates the preferred list of the subscriber’s SIM properly for the country.

For the sponsored network as the VPLMN, the HPLMN believes that the subscriber is trying to register to sponsor network actually. Besides this confusion, concluding the subscriber is at another country (Country Y instead of X) is a worse problem. There might be different settings for two countries and if the parameters are looser for the sponsor network’s country, this will cause problems for you.

The sponsor network might be a preferred partner and this prevents you to steer the subscriber to the preferred partners in country X. This may not be a problem if the sponsor network is “more preferred” than your preferred partners in the sponsored network’s country. But for the other case, this may mean that you lose some revenue.

Even though the sponsor network is more preferred than any preferred partner in the sponsored network’s country, you still should fix this case with configuration of the steering solution; to be able to retrieve correct reports grouped by countries.

Solution
Sponsor networks let the partners know about the sponsored network in general. They may publish the sponsorship as well the GT prefix for the VLRs used for that sponsored network. What you should do is to look for such publications from partner networks and alter the settings for your steering solution.

You will possibly need to create a new network to represent the sponsored network with the published GT range from the sponsor network's GT range. Considering your steering solutions allows that, then you should assign this new created network to the country / zone of the sponsored network.

With such setting, you will be able to correctly perform steering configuration for both sponsor and sponsored networks' countries. As roaming business is a considerable part of the revenues of the operators, having correct steering settings can be crucial to keep the costs minimized.

Trusting Google blindfolded

Or maybe I should have picked a title like “do not trust Google Maps picking a route for you”. It was my first time using Google Maps when I was driving between different spots during the vacation. It was always OK to use it in the city or at the main roads interconnecting cities. With this experience I had a trust using Google Maps on the smartphone so that I did not even need to follow road signs.

All rights reserved by jhellfire2000

Ending up in the field with pomegranate trees  following this route for Dalyan to Sarıgerme made me notice that Google may have erroneous data too.

Things started to go wrong at step 3, seeing that the road is not as straight as Google thinks so; but it was heading where Google promised to. I kept following and ended in the field. Too bad I was not driving a 4WD to follow Google's route comfortably.

It was not Google's mistake actually as Google is working with local providers for maps, like this map was provided by BaşarSoft. Checking the web I noticed I am not the only one having this problem.

This little experience made me notice there is still a lot to depend on online maps blindfolded.

Touristy spots of Istanbul

Coming across the World touristiness map and The Geotaggers' World Atlas by Eric Fischer online, I wondered Istanbul’s popularity map. While Eric Fischer’s map for Istanbul could give an idea about the popular spots, what it really provides is the mobility map of the photographers.

The different colors represent different modes of transportation: Black is walking (less than 7mph), Red is bicycling or equivalent speed (less than 19mph), Blue is motor vehicles on normal roads (less than 43mph); Green is freeways or rapid transit.

While world touristiness map provides the popularity info, the resolution is too low to check the popularity at the city level.

Inspired with both ideas I downloaded the locations of all 100K+ geotagged pictures taken around Istanbul at Flickr. Below is the heatmap produced with pictures since the beginning of 2010. While it would be faster to create small grids of the city and download only the number of pictures for each grid (like what "world touristiness map" did possibly), this would not be able to create a high precision map. Because of this, I tried to download all the picture coordinates.

It is no surprise that the most hot area is the historical part of the city and by Bosphorus. The spots on the sea are mostly on the route of ferry lines.

Knowing the city a bit, you could expect this result before trying to plot the pictures. But still if you do not know the city very well, it can teach you new spots close to the city to see like those small red dots at northern part.

Turkey's earthquake density map

Last year, while checking GoogleMaps' API I used Kandilli Observatory and EarthQuake Research Institute's recent earthquakes data to plot on the map.

For some time I was thinking to implement a heatmap layer for Google Maps, but could not have a chance to do so. Coming across Patrick Wied's code for the same purpose, I tried to play with his code a bit. Remembering the old earthquake map, I used Patrick's code to create a earthquake density map for Turkey using the Kandilli's data.



Above is a snapshot for the produced map. Colors changing blue to red implies increasing earthquake magnitude or more earthquakes around that area. Click the image to see the active version of the map.

Effect of network-based steering on international traffic

Network-based steering is one of the options to redirect the outbound roamers to the preferred partners and works by refusing registration messages from outbound roamers. By its nature, this causes some additional load on international links, which may mean additional cost for international connections. The amount of this additional load depends on several parameters such as the number of outbound roamers or preferred partner coverage performance. Though the exact value might be quite different for each network, an average of more than 5% additional load can be given as an educated estimate. This may mean that network-based steering may increase operators’ international connection costs.

SIM-based steering

Though it looks inevitable to have this additional load on the links, it is possible to reduce this overhead by employing network-based steering only when we have to. In other words, using adaptive SIM-based steering can reduce this overhead. SIM-based steering can be considered as the ultimate solution to steering. It is the best possible way of redirecting subscribers, making the handset work for the operator instead of trying to guess what is happening from the Home Public Land Mobile Network (HPLMN) side.

Using this approach, the handset is forced to pick the preferred partner as long as its service is available. It also makes sure that the handset switches back to the preferred partner even if it had to use an un-preferred partner because of coverage issues. It allows operators to have different levels of preferred partners, not just preferred and un-preferred partners. As operators can trust the handset now, knowing it will camp to a preferred partner if possible, it also allows subscribers to register and start using voice and data services more quickly. Removing the need for network-based steering also removes the overhead on international connections and increases steering performance.

However, most of the benefits of SIM-based steering depend on a specific command called Refresh. Without it, the settings will not be applied before the handset is restarted. Yet this command is not supported by every handset or SIM card profile. It means that the performance of SIM-based steering depends on the distribution of handset models as well as different SIM profiles on the network.

To maximize the benefits of SIM-based steering, it’s therefore vital to plan ahead while making SIM purchases to make sure that the new SIM profiles support this command even though there is no close future plan for SIM-based steering deployment. Even then, operators may still need to employ network-based steering for unsupported handsets as it is not possible to control the distribution of the handsets.

Roaming disabled subscribers

While network-based steering introduces an extra load on international links, it may need to be accepted in certain cases. Where it should definitely not be accepted is when it relates to steering outbound roamers who have their roaming disabled.

The percentage of roaming disabled subscribers who have their handset turned on while they are abroad is relatively high. The exact rate can be quite different for each network, but it can be around 30%. And worse, these handsets generate almost twice more registration messages per unit time, compared to regular subscribers. Based on these numbers, it’s clear that almost half of the registration requests are generated by roaming disabled subscribers. Applying network-based steering on these requests brings us nothing but pointless overhead traffic on international links.

Integrating and configuring the steering system properly is the solution to this problem. The steering system can access the operator’s internal database for this information or query the GSM nodes. Another solution can be to use passive mode solutions with steering on the return path.

While these solutions are to reduce the unnecessary cost of network-based steering, a much better option could be to turn those roaming disabled subscribers into roaming enabled, revenue generating ones.

See also Outbound Roamer Steering and Mobile Data Boost

Outbound Roamer Steering - Mobile Data Boost

The introduction of iPhone boosted the smartphone market as well as mobile data usage. Informa reports that by 2013 there will be one billion smartphones in use, thanks to iPhone and Google’s Android operating system. Each year, both the number of smartphones and their share in total number of handsets is increasing. Tablet computers (again triggered by Apple’s iPad) are another gadget pushing the mobile data usage a step further. According to Gartner, tablet sales will surpass 200 million by 2014.

Both smartphones and tablets consume a huge amount of mobile data, increasing the share of data revenues for operators. In parallel to this trend, Informa predicts that revenues from mobile data roaming will increase by a massive 246% between 2010 and 2015.

Within the context of the previous post, we have to pay special attention to data with our outbound roamer steering solutions.

Ensuring smartphones’ proper work on the network

As smartphones are becoming such an important part of the roaming business, we need to ensure that service is as seamless for them as it is for regular handsets. Unfortunately, the behavior of some smartphones can be different from regular phones as they are designed to be something more than just “phones”.

An example can be some specific models’ failure to interact with network-based steering solutions. Unlike the plain handsets they may accept having only data connection without any voice connection, resulting them to be stuck in a condition where the subscriber is not reachable for any calls or SMSs. Worse, this condition is not visible to the subscriber unless he tries to make a call. Although the solution to this problem can be as easy as enabling steering of GPRS messages, without such a care we’ll end up with unhappy customers and loss of potential revenues.

Segmentation

In the old times where the GSM traffic and revenues consisted of voice and SMS mainly, outbound roamers were tried to be redirected to cheaper voice partners. While this is still the case for most of the subscribers, with the increasing data traffic share we have to make sure data-heavy subscribers are redirected to the cheaper data partners. Steering systems segment the outbound roamers normally as prepaid, postpaid or VIP subscribers and treat them differently, but now we need a separate segment for data-heavy subscribers.

The question here is how to determine the data-heavy subscribers. Billing system integration is a proper option allowing the steering system to check the roaming history of the subscriber to determine if he’s a data- or voice-heavy subscriber. For the cases where the steering system does not support this kind of integration or operator cannot provide such an interface, checking the device database (IMEI) for the subscriber could be another option. Although this works on the assumption that all smartphone users are data-heavy subscribers, this is still better than doing nothing. And it can work properly for the terminals like data modems.

Some missed out aspects of outbound roamer steering

Roaming revenues have always been an important part of the operator’s income. The EU regulations introduced limitations of roaming charges both for wholesale and retail tariffs. The roaming revenues were expected not to be effected massively, with the idea that lowered prices would boost the usage. After several years with the regulations in practice, many operators argue that the usage did not increase as expected resulting in decreased roaming revenues. The regulations have also a water bed effect on the non-EU operators. To be able to compensate the revenues the EU operators may charge more to the non-EU ones which affects them too.

We hear a lot about the EU commissioner’s views on the roaming tariffs, which can be summarized as “no difference between the cost of making a call in one’s home market and across borders”. This is an indication that the roaming business will not get relief but will have lower margins each year. Reading news reports from non-European countries which ministers say the roaming charges are too high and should be lower, shows us that EU regulations are inspiring other countries too.

With these trends in the roaming business, operators are paying extra attention to the details to be able to maximize the profit while establishing subscriber stickiness. Outbound roamer steering has always been a way of maintaining higher profits from outbound roamers by redirecting them to the preferred roaming partners, though its nature is changing with the changing trends in the GSM world too. As a vendor with widespread roaming products serving to millions of subscribers and with close relations with operators, I would like to highlight a couple of issues about outbound roamer steering in the following posts.

Mobile apps gold rush

Ever impressed with the success of applications like Angry Birds for smartphones and it looked tempting working on your own application? Reading statistics like 12+ million paid copies and 30+ million free versions for a game which does not look quite complex might make anyone think and calculate about it ("12 million times that money and 70% of that ..."). It may even make you dream about becoming your own boss reading stories of people quitting their daytime jobs for iOS development.

Is it really that easy to be different and more successful than those more than 300.000+ applications? Considering thousands of publishers trying to achieve this, it is clear that it is not that easy.

Steve Costello writes about the driver of the success of Angry Birds:

According to Hed, the driver for the success of Angry Birds has been the “relentless social media buzz,” which has helped it stand-out from the raft of other mobile games products available for iOS and other platforms. “Getting featured in the app stores is important, but it does not guarantee automatic success. Working together with great brands helps, as does having a tremendous product. At the end of the day, being phenomenally successful puts you in the news and generates increasingly more interest and more publicity,” he says.

"The most successful developers tend to be time-served veterans, who can trace their heritage back to early Palm OS and/or Symbian titles" says Steve on another blog post. He also adds that to be able to successful you have to adapt yourself to put yourself in the best position for the future.

Perhaps there is a message for the developers currently working away either alone or as part of a small team, struggling to generate interest in a crowded market dominated by established giants. The most important thing is to keep going, and continually evolve to target whatever the market “sweet spot” may be. Admittedly, this is no guarantee of success – but it does seem to be the best way to maximise the opportunities.

How do you feel about trying your chance on your application after such reports?

Techno-junkies

Some time ago I watched some videos shot by the participants of Turkcell’s customer insight studies. The segment considered was the university students and they were given video cameras and asked to stop using mobile phones for some time. The cameras were for them to share their experience and thoughts.

Though I do not remember the videos in detail, I remember how they were complaining about the lack of communication. Even though some of the videos were a bit exaggerated, their message was clear and not surprising: Mobile phones became a huge part of our life.

A recent post at Engadget talking about a research by International Center for Media and the Public Agenda (ICMPA) reminded me that customer insight study. ICMPA’a research is sure more comprehensive but shares a similar idea: A class of 200 students living without any media (cell phone, iPod, television, car radio, magazines, newspapers and computer) for 24 hours and blog about their experience.

According to a new ICMPA study, most college students are not just unwilling, but functionally unable to be without their media links to the world.

The research study is not just showing that we’re incapable of living without media, but points out that we are actually addicted to it suffering withdrawal symptoms, just like real addicts.

Many students described their reactions to going without media for 24 hours in literally the same terms associated with drug and alcohol addictions: In withdrawal, Frantically craving, Very anxious, Extremely antsy, Miserable, Jittery, Crazy.

Do you think you could succeed (no texting, no Facebook or IM-ing) or could not resist to the desire and fail to complete assignment as some students did? Are you one of those junkies?