The telecommunications market we once knew has evolved significantly and will continue to do so as the lines between communications service providers (CSPs), managed service providers (MSPs) and content providers blur. CSPs are adding innovative services, technologies and content to re-invent themselves.
CSPs need to find ways to differentiate their services from competitors in today's crowded marketplace. A large part of this is ensuring strong performance and optimal quality of experience (QoE) for customers. At the same time, CSPs face industry commoditization with services' drop in price and fast-growing traffic leading to increased CAPEX requirements. Sometimes the tools in their back office are not able to keep up with the new services and technologies being launched and the associated challenges.
Earlier this summer I examined this dilemma as a panelist in Metanoia's online panel “Performance Management & Monitoring for Operational Efficiency.”
This panel consisted of executives from CSPs (Tata, Telus), over-the-top (OTT) operators (Twitter, LinkedIn), equipment providers (Cisco, ADVA) and software vendors (Packet Design, OPNET) and was attended by professionals from six continents, including more than 30 countries and 75 companies. We discussed various strategies including end-to-end performance monitoring and management, application performance and its relationship to network performance, and the relationship between OTT provider and infrastructure provider.
CSPs in attendance acknowledged that end-to-end performance monitoring and management is an important industry objective, despite technical and organizational challenges. Even the definition of an “end point” can vary upon the perspective, e.g. the end point for OTT providers could be a smartphone or an application running on a server, whereas the end point for an infrastructure provider might be the demarcation point of a Carrier Ethernet service.
Although end-to-end performance monitoring and management may present multiple challenges, many ways to ease this process were discussed. They are:
- Adopt all-IP technology in more of their network, e.g. all-IP mobile networks (LTE), next-generation Ethernet backhaul and unified communication for voice and video.
- Implement standardized services such as those defined by Metro Ethernet Forum (MEF). These services should leverage the embedded instrumentation such as Y.1731, 802.3ag that facilitate multi-vendor interoperability.
- Define service level agreements (SLAs) between infrastructure provider and the OTT operator and share the reports in a transparent, real-time manner that enables troubleshooting by either party. These SLAs should at minimum include latency, jitter and frame loss.
- Consolidate their OSS, enabling CSPs to economically scale their assets to manage the largest, most diverse and geographically dispersed next-generation networks, services and applications.
- Model the service to the underlying infrastructure to help measure the service quality and accelerate the troubleshooting of service degradation.
- Utilize service assurance performance platforms across all-IP mobile networks, to collectively manage their subscribers' services and quality expectations.
While application performance is key for OTT operators, many advanced CSPs are also introducing application-aware services such as application visibility, control and WAN optimization. Application performance not only depends on the application efficiency “code” and server capacity, but also the network that delivers it. Often times, when the application is slow, organizations tend to first troubleshoot the network. CSPs can better serve their customers by providing them with application visibility with reports such as top talkers, top applications and top sites, as well as by offering application-level SLAs that measure and report on the connection set up time and server response time. Investing in integrated application performance and network performance monitoring tools will also help CSPs to achieve consistency in customer facing reports and accelerate troubleshooting by mapping applications to infrastructure.
I was pleased to see that this panel received such a great turnout. Attendees walked away with best practices for application and network performance management that will enable their network to run efficiently and deliver the performance they need to ensure optimum service delivery to their customers.
