The European Commission’s ARECI Study1 highlighted that the resilience of public telecoms networks is not only important for public welfare, but is also vital to ensuring national economic prosperity. But who will bear the cost of measuring – and improving – resilience, and will national regulatory authorities (NRAs) across the EU know exactly what they are supposed to be measuring?
In the UK, the Digital Britain report states the Government’s proposal “…to give Ofcom a duty … to alert the Government to any significant deficiencies in the coverage, capability and resilience of the UK’s communications infrastructure…”.2 In order to fulfil this role of policing network resiliency, the responsible body will need to be able to obtain a valid number of network resiliency measurements and provide a meaningful view on the adequacy of national resilience. This means clearly defining:
- what is meant by ‘resilience’
- the method of measurement
- the level of resilience that is acceptable at a national level.
Agreeing a clear definition of network and – more importantly – service resilience is crucial in establishing a like-for-like, transparent, fair way to measure and report it. The method of measurement is equally important, and resilience cannot be measured directly. Network availability, that is, the duration that a network provides a service over a given period, is potentially a suitable measurement. However, measuring availability is not simple and only provides historical information.
Defining acceptable levels of network resilience is a complex area. Resilience is affected and supported by technology, network design, operational practices and market dynamics. Often commercial objectives and operational constraints mean that the minimum amount of investment is made in network resilience. There is a risk that increasingly large investments are made for an increasingly diminishing benefit. Typically, to inform the investment decision, a probability-impact assessment is undertaken, which considers the likelihood of a risk happening and the associated impact if it does. Investments resulting from such an evaluation may not be aligned with the wider requirements of the nation.
Calculating the cost
It also poses questions around the costs – and who should bear them – of measuring network resilience and increasing it if it falls short of national economic and welfare requirements.
Measurements need to be comprehensive and statistically significant. This may mean investing in many measurement devices in many networks. The cost of measuring resilience and the potential cost of providing the required resilience could be very significant. This may be especially challenging for new entrants – potentially increasing market entry barriers – and smaller established service providers. There could be difficult trade-offs between wider public policy objectives and more specific regulatory policy initiatives aimed at supporting competition.
Resilience is especially complex in the case of next-generation networks that provide multiple services with different quality of service demands, using the same physical infrastructure. Does it really matter if some services are not available? A single network connection that failed for five minutes during which an emergency call was attempted is significantly more important than thousands of subscribers not being able to access Twitter for several hours. Any incremental cost associated with making networks acceptably resilient for life-line services is obviously less difficult to justify than for other services – but can you physically separate life-line services from others?
Prevention not cure
Analysys Mason would also recommend that prevention is better than cure. There is little point in reporting and acting on a network that has failed after it has already caused disruption to welfare and economic systems. Preventative measures embodied in network design principles and operational processes are critical factors to be considered. This implies that a set of network design and operational process guidelines should be defined and agreed by all key stakeholders including responsible bodies e.g. NRAs and network operators.
NRAs may soon be called on to police resilience, and operators may be forced to take – potentially costly – action to provide improved service levels. Analysys Mason has a deep knowledge of what drives regulatory policy and operators’ commercial costs and benefits. We provide advice to NRAs to help them make informed policy decisions. We also provide advice to network operators regarding commercially viable solutions that could help them meet their commercial objectives and the potential national security and welfare obligations that seem to be on the horizon.
For more information, please contact Mike Pearson, Partner at mike.pearson@analysysmason.com.
1 See Availability and Robustness of Electronic Communication Infrastructures (ARECI), Alcatel-Lucent for the European Commission (March 2007)
2 Digital Britain, Final Report (June 2009)