As fibre becomes more central to broadband, cloud and enterprise connectivity, the resilience of digital infrastructure is increasingly shaped by the physical routes carrying traffic between network locations. Network route diversity is becoming more important because additional connectivity does not necessarily create an independent alternative when multiple services follow the same corridor, duct, landing station or aggregation site. A single physical disruption can therefore affect several services that appear to be separately provisioned. The resilience question is consequently shifting from how much connectivity exists to how many genuinely independent paths are available across the network.
The scale of international dependence makes this issue more significant. The International Telecommunication Union estimates that submarine cables carry more than 99% of international data traffic worldwide and recorded more than 170 cable repairs globally during 2025. The commercial submarine cable network now extends beyond 1.7 million kilometres. These figures illustrate why failures affecting individual cable systems can have consequences beyond a single operator or market, particularly where international traffic is concentrated through limited physical corridors.
Geographic Diversity Extending Across the Connectivity Chain
The need for network route diversity extends across terrestrial fibre, metropolitan networks, cable landing stations and data centre connections rather than stopping at the submarine segment. Two operators may provide separate services while still relying on the same underlying infrastructure, creating a shared failure point. Resilience therefore depends on the physical independence of the routes as well as the number of providers supplying connectivity.
This principle is increasingly influencing network expansion projects. Google’s network architecture incorporates diverse and disjointed routes, while AWS uses multiple network pathways and topology reviews to identify potential points of failure. New international systems are also being designed around geographic alternatives. Google’s America India Connect includes multiple subsea paths and fibre routes, while TalayLink between Australia and Thailand is intended to establish a geographically different path from existing regional corridors. Similar diversification projects are developing across the Middle East, South Asia, Africa and Europe.
Network route diversity is therefore becoming part of network planning rather than a separate contingency measure added after deployment. As fibre carries a larger share of critical digital traffic, operators and infrastructure providers increasingly need to consider whether new routes offer genuine physical separation, how traffic can be moved when a path fails, and whether sufficient alternative capacity exists to maintain services during disruption.
New fibre routes are increasingly being designed to add geographic alternatives, helping reduce dependence on concentrated physical corridors.
Rerouting and Recovery Extending Resilience Beyond Infrastructure
Building an independent route is only one part of resilience. Networks also need sufficient spare capacity and traffic-management systems to move services when a primary path fails. Google describes using diverse network paths together with automated rerouting to shift traffic around disruptions, while AWS maintains multiple pathways and reviews network topology to identify potential points of failure. These approaches show that network route diversity works most effectively when physical alternatives are combined with routing and operational mechanisms that can use them during an outage.
Recovery capacity is equally important. The ITU has identified repair readiness, vessel availability and regulatory coordination as important factors in restoring damaged submarine infrastructure. Network route diversity can reduce the immediate impact of a failure by providing alternative paths, but it does not remove the need to repair the damaged asset. Operators therefore need to consider route independence, spare capacity, monitoring and restoration procedures together when designing resilient fibre infrastructure.