TPRC 2024: The 52nd Research Conference on Communications, Information and Inter

Link to paper

Abstract:
This study investigates the importance of broadband internet access reliability in underserved Alaskan communities. Analyzing measures of network latency and outages collected from anchor institutions over a two-year period, we demonstrate the inadequacy of existing broadband definitions to ensure adequate quality of service (QoS). Broadband reliability measurements are comprehensively assessed by identifying overall trends in network performance, the presence of cyclicality, and the impact of natural experiments.

Using quasi-experimental designs, the study models the impact of significant incidents on broadband service reliability, such as the Quintillion submarine cable getting cut by sea ice in the Summer 2023. Fourier transformations are applied to identify cyclical trends in broadband reliability, revealing temporal variations and underlying patterns influencing service quality over time. Additionally, the research assesses QoS in ISPs and regions, facilitating a comparative analysis of performance metrics and identifying changes in service quality over the study period.

Alaska represents an edge case, but the empirical evidence of these challenges and the implications of inconsistent broadband reliability has broad application to underserved rural regions across the United States. These insights are relevant to policymakers, stakeholders, and ISPs as they prioritize once in a generation broadband infrastructure investments. This research underscores the critical role of reliable broadband access in supporting critical infrastructure, promoting socio-economic development, and addressing the digital divide.

Link to paper

Abstract:
Focused on broadband service accessibility in rural/remote communities, this study investigates mobile Internet access and usage in the Northwest Territories (NWT) of Canada. Using primary data collected from Northern mobile phone subscribers, the study explores mobile phone ownership, usage frequency, and hotspotting behaviour in hub and peripheral communities. Demonstrating variations between two types of rural/remote communities, the study found significant differences in mobile phone access and use. Residents in peripheral communities showed higher mobile phone ownership and greater reliance on mobile Internet, averaging 4.2 hours of daily usage compared to 3.6 hours in hub communities. These factors are attributed to the lack of adequate household connectivity options and the necessity for reliable, on-the-go Internet access. Additionally, the study highlights motivations for mobile Internet usage, such as convenience, better speed, and broader coverage, with many participants citing dissatisfaction with existing household Internet services.

The findings also reveal a high incidence of mobile hotspotting behaviour, particularly in peripheral communities. This indicates a propensity among Northern residents to share mobile Internet connectivity across devices and end users. While mobile service costs were marginally higher in peripheral communities, this difference was not statistically significant. However, household service plans in peripheral communities include data restrictions and overage fees, which suggests the higher instances of hotspotting behaviour and adoption of mobile services may be a way for people to manage such costs. In the context of recent policy and industry developments concerning support for mobile infrastructure and service deployments in rural/remote and Northern areas, these insights contribute to understanding the role – and limitations – of using mobile Internet to bridge digital divides. While mobile services could offer a cost-effective solution to improving digital access, they also reflect limitations related to device capability and quality of access and may exacerbate the so-called “device divide”.

Link to paper

Abstract:
For decades, the policy priority to bridge the  digital divide in the United States was to obtain additional funding to build out and upgrade networks, particularly in rural and disadvantaged areas. However, in the past two years, numerous federal programs, plus some state and other initiatives, have allocated billions of dollars to achieve the goal of “universal broadband.”
While resources are more limited in Canada, funding is available from some federal agencies as well as some provincial and other sources. The emphasis in both countries as been on funding for capital projects such as provision of optical fiber for both transport and local networks or equipment for fixed wireless broadband networks. However, there has been little attention to the sustainability of these networks such as business models, and management and technical skills to continue their operation.

While research on the impact and sustainability of recent federally funded projects will take time, this paper examines the challenges they may face and strategies to overcome or minimize them. The following issues are addressed:
• Engagement: consultation, training and hiring
• Participation: partnerships and ownership
• Sustainability: Financial, technical and resilience
• Competition and disruption.
Engagement with local or Tribal governments can not only explain the proposed project, but also identify issues such as access to existing facilities, rights-of-way, services to be provided and pricing. While the FCC now requires Tribal consultation for projects on Tribal lands, preliminary research indicates that there is little enforcement of this requirement. Federal infrastructure programs now mandate engagement with local or Indigenous governments, but it is unclear how compliance will be formalized and monitored. Engagement can also identify opportunities to train and hire local residents for both network installation and ongoing operations and maintenance. Consultation may also identify needs for digital literacy training so that households and organizations can benefit from the infrastructure investment, contributing to local economies.

The paper reviews various models of participation including partnerships and ownership, and examines the implications of programs that fund some Indigenous organizations with little management experience or technical expertise, and must rely on contractors.

Sustainability can include financial viability as well as technical sustainability. Redundancy can be critical for sustainability to ensure continuity of service, for example, satellite terminals to back up fiber networks, and ring transport architecture to provider alternative pathways. The paper provides evidence of extensive outages where projects did not include funding or consideration for redundancy.

Technical sustainability can also include future-proofing to prevent short-term obsolescence at minimal cost. Requirements for inclusion of dark fiber and upgradeable switching can extend useable project life. The paper provides examples of relatively recently installed rural networks that could not meet increased demand during the pandemic, and unforeseen costs incurred to upgrade switching, such as for 988 access.

Concerning competition, some incumbent providers state that allowing competition would discourage investment in regions with low populations, and conversely, that if competition is viable, it should only be facilities-based. However, the paper provides evidence from Alaska and Canada of open access competition, as well as facilities-based competition, most recently from Starlink.

The paper concludes with questions to be addressed, and the need for independent evaluation to determine the impacts and lessons from investments of billions of dollars in rural broadband in Alaska and northern Canada.