GSM spectrum holes: another type of "white spaces"

Another frequency band just starting to raise interest among the research community is the Global System for Mobile Communication (GSM) bands, namely, 890-960 MHz and 1710-1880 MHz. The GSM standard is among the largest mobile communication technology worldwide. However, while looking to the unique subscriber penetration (and not SIM penetration which is usually confusing and irrelevant for such contexts) of mobile services in Sub-Saharan Africa for example, the rate is no more than 31% at the second quarter of 2013 (source: GSMA [1]). This is because rural and remote areas in Africa still have poor GSM coverage and undergo a lack of GSM deployments.  As a result, GSM bands are underserved in these places and consequently can be exploited and accessed in the objective of providing GSM capacity expansion to rural population. In analogy to TV white spaces, the GSM traffic possibly leaves spectrum rooms called GSM white spaces.

Similar to TV bands, GSM wavelengths are longer and so can travel farther and penetrate obstacles much better.  GSM white spaces do not need a geographic coordination through a central database since GSM signals already implement robust interference mitigation mechanisms.

More importantly, two experimental trials have been conducted to prove the concept of cheaper GSM service or the so-called Community Cellular Networks (CCN).  In a small village called Villa Talea de Castro, in Oaxaca, Mexico [2], indigenous communities, civil organizations and universities have setup an autonomous and locally-managed GSM infrastructure to enable phone calls and texting for population in this village more than ten times cheaper compared to traditional telecom operators’ prices. Although current rulemakings didn't allow any use of GSM bands in spectrum sharing contexts, this village have benefited from some provisions in the Mexican constitution, allowing for some exceptions in case of social purposes, to circumvent these difficulties. It was amazing to see such isolated villages getting connected using GSM handsets [3]. The second success story comes from Papua in Indonesia [4] where some researchers from UC Berkeley have built a prototype for low-cost GSM macro-cell.  Their idea was smart and it consists of intelligently gathering regular measurements reported by mobile phones, concerning their associated neighbor lists, into a central database which will be used later to monitor the used frequencies on the covered area. The project has used OpenBTS [5] regarding its flexibility and cost-effectiveness.

Lastly, I think both TV and GSM white spaces have a great potential to be more attractive to rural and remote communities to self operate their own and affordable next-generation networks.

BIBLIOGRAPHY

[1] http://www.gsmamobileeconomyafrica.com/Sub-Saharan%20Africa_ME_Report_English_2013.pdf.

[2] http://timesofindia.indiatimes.com/world/rest-of-world/Ignored-by-big-companies-Mexican-village-creates-its-own-mobile-service/articleshow/22094736.cms?referral=PM.

[3] http://edition.cnn.com/video/data/2.0/video/bestoftv/2013/12/10/mexico-mobile-phone-cell-service-rafael-romo.cnn.html.

[4] http://www.eecs.berkeley.edu/Pubs/TechRpts/2013/EECS-2013-198.pdf.

[5] OpenBTS. http://openbts.org.