Making the Autonomous Network a Reality

It is well known that we have become a more connected and on-demand world. Gartner forecasts that 8.4 billion connected things will be in use worldwide in 2017, up 31% from 2016, and will reach 20.4 billion ...

Making the Autonomous Network a Reality
Making the Autonomous Network a Reality

It is well known that we have become a more connected and on-demand world. Gartner forecasts that 8.4 billion connected things will be in use worldwide in 2017, up 31% from 2016, and will reach 20.4 billion by 2020. To keep pace with user demands for applications like video streaming, Internet of Things (IoT), mobile broadband and cloud-based services, network operators know that using a static, one-size-fits-all model is no longer sustainable. Much like the future of cars, communication networks must become more autonomous and self-driving to meet fast growing customer demand.

Why the status quo no longer works

Before we jump into what needs to change, it's important to understand why continuing with the present mode of operation (PMO) won't meet the networking challenges of the next decade. Today, optical networks are engineered to operate in a static mode, using best-guess predictions of worst-case capacity demands and system conditions. If capacity is overestimated, network providers are left with an inefficient network, paying for equipment that is sitting idle. On the flip side, if new demands arise unexpectedly, they need to go through new lengthy and costly planning and deployment cycles.

While the PMO works when capacity demand is relatively predictable, it is simply impossible to plan for future demand in today's environment - especially as maximum demand requirements are growing exponentially. Networks of tomorrow must not only scale to meet massive capacity growth, but also need to be more agile and programmable to handle the increasingly unpredictable nature of traffic requirements.

Transform margin into capacity

Critical for modern networks is having the necessary features to make it possible for operators to tune, control and dynamically adjust optical capacity. Today, the only way capacity can be turned up and down is to pre-deploy hardware and reserve network bandwidth in advance. Moving forward, network operators need the right technology in place so they know the amount of available margin in the network, and then can automatically upshift coherent optical capacity where there is enough margin to do so.

To make this concept of an autonomous network a reality, operators need to take tangible steps today as the shift won't happen overnight. There are a series of next-generation optical and software technologies that need to be in place before operators can extract more value out of existing network resources. For example, networks will require:

  • Variable bitrate coherent optics: A key element in enabling on-demand capacity, variable bitrate and software-configurable coherent optics can match optimal channel capacity to available system margin for a specific network path.
  • Efficient client-line matching of throughput: The ability to efficiently map a flexible number of client signals to the variable line capacity is needed. Centralized optical transport networking (OTN) or packet switching is an efficient architecture for this purpose. A high-density muxponder solution is another viable option. In the future, operators can also leverage impending flexible client protocols such as Flex Ethernet.
  • A flexible grid, agile photonic architecture: This type of architecture provides the ability to reroute channels of variable spectral occupancy across any path and across any optical spectrum in the network.
  • High performance streaming telemetry to gather network data in real time: This data can then be used by software applications to provide a deeper understanding of network behavior in real time, which in turn allows it to be tuned for optimal performance in all conditions.
  • Multi-layer software: This software must have the ability to measure system margin at any point in time, perform link engineering for different what-if scenarios and per-channel optimization, push the right provisioning parameters into the network, and coordinate between the multiple layers for appropriate handling of services.

Soon, working to create an autonomous network through real-time monitoring and analytics will not be a luxury, but rather a requirement as it will directly impact operators' ability to achieve their business goals. Having a network that can meet demand without manual intervention, operators can accelerate time to market when faced with unpredictable service demands, improve service availability during a disaster recovery, and eventually, dynamically move bandwidth easily across the network to address mobile capacity requirements.

Helen Xenos is director of portfolio marketing at Ciena.

More in FTTH