Reliability of Rural Broadband Networks
By Fletcher Kittredge:
Let’s talk about network reliability. In rural America, we have been reluctant participants in an experiment to see if modern society is a network dependent society. Some of our communities lack any broadband at all, and others have broadband with spotty coverage and/or poor reliability. The results of this experiment show that while broadband alone is not sufficient for creating a vibrant society, towns that lack broadband lose economic viability and population. Neither young people or businesses will move to or stay in a town without broadband. Today good network infrastructure is a necessity and even more so in the future.
Broadband infrastructure’s key components are speed, reliability, and security. Broadband needs to be good enough for residents to:
Have a wide range of leisure activities
Be socially and civically engaged
Have educational and health options
Have economic opportunity
Given the profile of the rural economy, it is a requirement that residential broadband be of the same quality as business broadband. Telecommuting, home businesses, and farms together are a very significant and growing portion of rural economic activity and all of these depend on broadband to the home.
An unstoppable trend is the increasing importance of telehealth. We are in the early days of the telehealth revolution but it is likely that in the future that life-critical healthcare will extend into the home requiring a higher standard of reliability.
In terms of reliability, for most people the “business in the home” aspect creates the strictest reliability requirement (though parents of school age children with homework may disagree!) People can live temporarily without streaming and social media but their employers and customers will not tolerate them not being available. In practical terms, broadband connections must be available (up) 99.97% of the time 365 days x 24 hours per year. This works out to 19 hours of downtime per year.
Implications for rural ISPs
By far the most cost effective way to reach reliability goals is to design it in during construction. The good news is that all of rural America will have fiber networks in the next decade, but for most of rural America the network remains to be built. The two bits of bad news are first that some ISPs are building networks today not for future needs, or even today’s needs, but how networks were used in the past. They design as if the only residential use of the ‘net is for entertainment. Second, global warming is making it hard to predict future natural disasters. Frequency of violent storms may increase and areas that haven’t flooded in hundreds of years can be flooded.
Designing Networks for Reliability
The most frequent cause of network outages is loss of power. This is particularly true for cable and wireless networks. Fiber networks are less susceptible both because they use significantly less power and because fiber transmissions can go longer distances meaning there are less points to be powered. For extended outages, both batteries and generators are necessary. Cable networks and wireless networks frequently generators for major distribution points and business parts, but often only have batteries for points deep in residential neighborhoods. Because 5G is dependent on small cells for higher speeds and it is impractical to have generator backup for small cells, 5G may be less reliable than earlier protocols.
Fiber networks don’t need as many facilities as DSL, cable, or wireless networks but they still need some. Facility siting criteria needs to recognize the rapid change in climate. It is not sufficient to be above the 100 year floodplain; the 1,000 year level should be used instead.
After power, important item for reliability is path redundancy for fiber. It is common to have path redundancy for “middle mile” fiber; the middle mile is usually built in rings. It is much less common to design the last mile in rings, but as much as possible it should be done so. Path redundancy needs to be driven deep into the network.
Implications for Government
The most fundamental and important role of government is to accurately and competitively neutrally collect all the necessary data to measure reliability and then make it timely available in a transparent and easy to use manner. For decades, state and Federal government have measured the reliability of power and telephone networks and these efforts should extend to the broadband networks which society is dependant on. The FCC is constantly improving its efforts to measure the speed and availability of broadband. Those measurements should be incrementally extended to measure reliability as well.
Governmental definitions of broadband should include a reliability component. It is not “speed”, it is “speed/reliability”. Requirements for governmental support of broadband infrastructure should have reliability criteria.
Implications for the Maine Broadband Coalition
The Maine Broadband Coalition (MBC) has been rightly focused on extending broadband to all communities of Maine. The primary measure has been speed. As more fiber networks get built out to rural areas, it is important that the goals broaden to meet the real needs of Maine communities. We need a network that can attract and retain population and build a vibrant economy for the 21st Century. To meet that goal, the network must be reliable and MBC
should include reliability in its goals and advocate for the same.
A few ways in which reliability can manifest in Maine:
Prioritizing network enhancement local ordinances to enable path redundancy
Facility citing amidst climate vulnerability assessments
Integrating micro-grid development and last mile ring design
Supporting municipal pole attachment rules