MPT 1 Ooredoo 1 Myanmar Telenor 1 Myanmar Data Plan 1.35 GB Internet Monthly 1.35 GB Smart Internet Month Pack 1.25 GB Price MMK 6500 MMK 6850 MMK 6600 Advertised Speed Not advertised Not advertised Up to 2 Mbps Validity 30 days 30 days 30 days Table 6: Plans selected from the different operators for QoSE testing There are many tools and methods that can be used to measure broadband Quality of Service. Often, the underlying methodology adopted is one of the reasons as to why readings vastly differ based on the tool used for 15 testing. LIRNEasia considered Ookla’s SpeedTest and the Open Technology Institute’s Network Diagnostic Tool (NDT). The methodology of the former is partially opaque whereas NDT is well documented (The methodology and technical details of the NDT can be found at https://github.com/ndt-project/ndt/wiki/NDTTestMethodology). The researchers conducted diagnostics on two week days and a weekend (four days in total) during the second week in May 2016, at multiple times of the day (8:00, 11:00 15:00, 18:00 20:00 and 23:00) to iron out potential peak / off-peak traffic variations. Researchers covered locations in all the states and regions. The testing locations can be seen in figure x. Ideally, multiple locations would have been covered to geographically represent a state/region. However, as this was a resource-intensive exercise, one township was selected per state / region. Diagnostics were carried out in one location per township. The intended use of the server in Bangkok was to replicate it as a locally hosted server. This was the initial hypothesis. However, on completion of the diagnostics it was evident that the NDT server in neighboring Bangkok cannot be a proxy for a server within the national domain. Trace routes from Yangon to www.google.com and to the NDT servers in LA and Bangkok illustrate the routes taken between the client and server. Several private IP addresses were encountered which are most likely to be within the service provider’s network and packets reached / was routed through Bangkok after a significant number of hops. Ideally, a locally hosted server would be used to measure broadband QoSE within the local domain. LIRNEasia is in the process of evaluating three other mobile apps and will employ a crowd-sourced approach to increase the geographical coverage of measuring as broadband quality, which is highly location sensitive. 4.2 Capacity building 4.2.1 Capacity building for policymakers At the initial states of project planning, it was envisaged that one course on e-government and one-course for legislators on telecommunications for regulators would be carried out. No courses took place in the first year of 15 Zuhyle, S. & Mirandilla-Santos, G. 2015. Measuring Broadband Performance: Lessons Learnt, Challenges Faced. Available at: http://www.cprsouth.org/wp-content/uploads/2015/08/CPR-south-2015_MeasuringBroadband-Performance-Lessons-Learnt-Challenges-Faced_PP21_Final.pdf and Bauer, S. et al. 2010. Understanding Broadband Speed Measurements. Massachusetts Institute of Technology. Available at: http://groups.csail.mit.edu/ana/Publications/Understanding_broadband_speed_measurements_bauer_clark_leh r_TPRC_2010.pdf