Telcos must automate or risk survival, says Kiran Inampudi, Solutions & Go-to-market Leader, Cisco. At TM Forum Live! Asia next month in Singapore, he will discuss how to use data science for effective network operations.
Most telcos worldwide have started to experience an alarming trend: Revenue-per-bit is falling faster than the cost-per-bit – primarily driven by factors like competition, pricing, regulation, etc. This revenue/cost crossover is driving aggressive transformation. Telcos recognize it is impossible to meet bandwidth requirements of the future by operating the way they have in the past.
Transformation initiatives like software-defined networking (SDN) and network functions virtualization (NFV) have had limited impact so far and have failed to improve the efficiency of operations or reduce operating expenses. Most industry experts believe that telcos face a stark new reality: automate or die!
The path towards an automated future is not going to be easy, and it’s likely to be a journey that’s phased over many years. Telcos are stuck with managing hybrid environments involving both legacy and modern technologies.
Adopting automation is a broad proposition, requiring investment in technology, people and process change. Telcos need to start thinking like software companies, and this often requires a change in mindset.
The path to automation is data-driven
Self-driving cars provide insight into the path that telco automation is likely to follow – collecting massive amounts of data, allowing algorithms to navigate their way through routine tasks, and implementing self-learning systems that can adapt to unpredictable situations. The result is likely to be smart network management software that can perform many telco operations tasks with good reliability.
Data science has the potential to transform network operations, including reducing the heavy load of manual effort involved in network monitoring, troubleshooting and optimization.
The benefits of data science – a real-life scenario:
A large telco has experienced a significant outage in their core network caused by a software bug. Their current network management systems and operations teams could not detect the issue for several hours. The root-cause analysis showed that the control plane and data plane were out of sync, causing the traffic data to disappear into a black hole for several hours.
To ensure this did not happen again, the telco decided it needed to collect telemetry data in real time to be able to detect such issues quickly. It did so by using modern network devices which can stream data continuously, and it also leveraged data science techniques such as unsupervised learning (a type of machine learning algorithm) to correlate the data, identify patterns and detect anomalies proactively in real time.
Such proactive approaches to network operations are becoming a powerful means for making networks more reliable and secure.
Book your place now for TM Forum Live! Asia in Singapore this year. Whether you are interested in reducing customer churn, delivering new digital services or virtualizing your network, the event offers content you can use and inspires innovation.