Estimating the effects of Internet Exchange Points on Fixed-broadband speed and latency

Estimating the effects of Internet Exchange Points on Fixed-broadband speed and latency

Date: 
Monday, December 9, 2019
Type: 
Public information and advocacy materials
Abstract

Access to the Internet contributes to socio-economic development. However, access to affordable and reliable broadband connectivity is not universal, particularly in countries with special needs (Least Developed Countries (LDCs), Land-locked Developing Countries (LLDCs) and Small Island Developing States (SIDS)). According to the latest ICT statistics from the Economist Intelligence Unit’s ‘Inclusive Internet Index 2019’ , the average fixed-broadband access, speed, latency, and affordability, are well-developed in high-income countries, but significantly lacking behind in low-income countries.

Internet exchange points (IXPs) are physical (switch or server) locations where different networks connect to exchange Internet traffic. IXPs are vital for improving the affordability and quality of broadband connectivity, within and between countries, by keeping Internet traffic ‘local’ within the network of connected networks, thereby keeping costs and latency low.

This Working Paper estimates the relationship between the number of IXPs and fixed-broadband speed and latency in 85 countries from 2016 to 2018, using a balanced panel dataset developed by the Economist Intelligence Unit for its ‘Inclusive Internet Index’. While several studies have established a positive role of IXPs on Internet speed and latency, majority of the earlier studies are technical studies examining the traffic routes in specific networks. This Working Paper therefore contributes to this literature by triangulating earlier findings using an econometric model. The recent availability of the panel dataset on IXPs, speed and latency by the Economist’s Intelligence Unit has made this exercise possible.

The preliminary findings highlighted a statistically significant and positive relationship between the number of IXPs and fixed-broadband speed. For every 1% increase in the number of IXPs per 10 million inhabitants, the speed of fixed-broadband download speed (Kbps) is expected to increase by about 0.14%. The preliminary findings further highlighted that the presence of IXPs are driven by income with higher number of IXPs in high-income countries and vice versa. In addition, countries with higher number of IXPs also tend to have higher access to broadband Internet and more affordable.

Despite the benefits of IXPs, challenges remain on establishing IXPs. In particular, the need for: collaboration and building trust between several stakeholders (national and international); neutral location and management of IXP as a platform for all operators to connect; conducive regulatory environment that supports an open market for telecommunication services. As a result, these challenges pose important policy implications for policymakers on ensuring the sustainability of IXPs.

Download: 

Estimating the effects of Internet Exchange Points on Fixed-broadband speed and latencyDownload