The National Telecommunications Commission is close to issuing a memorandum order that will update a previous issuance on minimum speed of broadband connections. But this time, the policy is targeted at measuring internet quality of service over setting minimum standards.
The order is the result of months of public consultations and technical working group meetings attended by big and small telcos, consumer groups and think tanks, and concerned government agencies such as the departments of justice, and trade and industry.
If and when NTC issues this memo order on broadband QoS, it is hoped that Philippine internet would improve from its abysmal state, ranking only second to war-torn Afghanistan in average download speed. The QoS measurement results might also help give a clearer picture of the problem areas that need to be addressed by government, service providers, and consumers.
The initial working draft was based on the proposal of Democracy.net.ph, formed by a group of concerned Filipino netizens who lobbied against data capping back in 2011. The Netdems, or founders of Democracy.net.ph, proposes to set minimum standards for data rate reliability, service reliability, and overall reliability, as well as “best efforts” offers. It also wants to ensure that consumers are properly informed by ISPs of their service offers and to establish rules for refunds and rebates in case of subpar service.
The draft now contains recommendations from LIRNEasia, a regional ICT policy think tank, who has been conducting broadband QoS experience studies in the Philippines and other Asian countries since 2011. The think tank proposes that NTC officially identify the diagnostic tool that will measure ISP performance metrics, which inclue throughput (download and upload speeds), latency, packet loss, and jitter. It also suggests that local ISPs publish their average speeds per location (e.g., city level) to allow consumers to make an informed choice. LIRNEasia found that Philippine ISPs offered the lowest value for money in terms of cost and actual download speed in tests conducted in Q1 2014.
In at least two public hearings, the two dominant telcos, PLDT and Globe, recommend that the NTC measure and monitor only the throughput and data volume, as these are the only metrics promised to subscribers. They also propose that internet QoS tests be done only within the ISP’s network where the variables are within their control.
However, measuring QoS from within the network may not necessarily reflect the actual service experienced by the end user. In the Philippines, for example, where a lot of websites are hosted abroad, it would be relevant to measure latency or round-trip-time vis-à-vis an international server. In countries that use probes (e.g., RIPE Atlas) or where the regulator provides testing equipment (e.g., SamKnows) or software (e.g., M-Lab’s network diagnostic tool) to volunteers, the tests are done from the end users’ location. This approach will most likely gauge the actual quality of connection in the last mile.
The importance of a sound methodology can not be overemphasized. A large dataset should be able to offset the effects of the uncontrolled variables and anomalies in the internet QoS tests. LIRNEasia proposes that the NTC-mandated diagnostic tool also be made available to the public, so that a wider area and population is tested.
The Private Association of Philippine Telephone Companies (PAPTELCO) representing the small telcos raises a valid concern about countryside resellers, or those who rely on the infrastructure of the big telcos, whose business might get negatively affected by the tests and publication of results. It urges the NTC to look into the lack of capacity, especially in remote rural areas, which would affect the performance of small providers.
Ideally, all service providers should comply with some form of QoS standards. However, there are ways to take into consideration those who serve a small subscriber base or missionary areas. In some countries, like Singapore, broadband QoS metrics for compliance are imposed on the large ISPs only, or those who serve at least 10% of the overall subscriber base. In the Philippines’ case, the urban centers might be a good place to start the internet QoS measurement where the big players operate.
This is a crucial juncture for NTC, as with many other regulators worldwide who are struggling with broadband policies, which include striking a balance between encouraging investment to expand access and maintaining quality of service for consumers. The proposals from the service providers and advocacy groups both have merits, and their recommendations have been discussed exhaustively over the past seven months. However the NTC decides, one thing is clear: this policy is long overdue, and it could only serve to improve Philippine internet.