Mobile Spectrum Requirement Estimates: Getting the Inputs Right

Executive Summary LS telcom provides spectrum management services to regulators, operators and spectrum users around the world. In addition to software and hardware solutions, LS provides consultancy services in spectrum management. A common request amongst regulators is to provide an evaluation of the amount of spectrum required for commercial terrestrial mobile broadband (IMT) services. In trying to assess the demand for spectrum for IMT services, we have examined the model used by many regulators around the world to produce the forecasts which will eventually be used to inform decisions at World Radiocommunication Conferences (WRC). Using this model, developed by the International Telecommunications Union (ITU), we have identified a number of issues which mean that the model, as it stands, does not provide useable nor useful results. We are not alone in our findings. Several parties (including the Russian spectrum regulator and the European Broadcasting Union) have questioned the values produced for IMT spectrum demand generated by the ITU model. Telecommunications analyst and consultant Tim Farrar of TMF Associates had also identified flaws in the model. In order to try and inform a discussion of how the model can be modified, so that it is useable for predicting IMT spectrum demand, LS telcom and TMF Associates have jointly produced this report to share our findings. 1 present forecasts for growth in the total amount of mobile traffic in the world to 2020 and then models spectrum demand based on traffic density in across a variety of service environments (SE) which include urban, suburban and rural areas. It concludes that the demand for spectrum in 2020 is between 1340 MHz and 1960 MHz (in low and high demand situations respectively). The ITU model purports to have used, amongst others, the UMTS Forum traffic forecasts as a basis for developing the spectrum demands. However, the traffic forecasts in the ITU model appear to be built-up from different sets of assumptions, which are at odds with the UMTS Forum forecasts. A comparison of the traffic density figures actually used in the ITU model with figures based on the UMTS Forum forecasts for urban and suburban areas show that the traffic density figures in the ITU model are at least two orders of magnitude too high. We have identified that there is a fundamental problem with the ITU model in that the traffic density does not appear to have been benchmarked against total predicted traffic …