Sunday, December 11, 2022

Blueprint: Why CSPs need to shift to support root-cause analyses

by Stephanie Sculley, Head of Analytics Practice, Advanced Consulting Services West at Nokia

Today’s networks are complex to say the least. Because of that, it can be difficult to determine the underlying cause of an issue that’s hindering the customer experience. 

For some communications service providers (CSPs), chasing the wrong suspect can result in wasted time and effort while the true criminal runs free continuing to frustrate customers. 

This unproductive cycle is prevented with root-cause analysis, which identifies the actions that are needed to solve problems, minimizes churn and keeps customers happy, or at least happier. Not only does this save time, but it also helps CSPs keep their decisions customer-focused.

Nonetheless, there are three areas where CSPs often get stuck when trying to conduct root-cause analysis to address these problems, but fortunately, switching up one’s mindset can help to mitigate these common errors. Let’s find out how.

You Don’t Know If You Don’t Ask…a Lot: How to Solve the Problem with Meaningful Questions 

A cold has similar symptoms as an allergy, the flu or even a more serious diagnosis. So, if someone says they have a cough or a runny nose, the doctor needs to ask the right questions to come to an accurate diagnosis. This may even involve doing research or even consulting other professionals, who may have experience in other areas.  

While not the exact same scenario, this is a similar situation to what CSPs face when trying to determine the cause of network issues. Performance and service-related issues can originate from various sources from device manufacturers, content or app providers, user behavior, device manufacturers, or even the network itself, making it near impossible to figure out the cause of problems on one’s own.

Thus, it’s critical that CSPs ask the right questions to determine the cause of lagging key performance indicators (KPIs), growing customer complaints and churn.

Finding a true root cause can be like peeling an onion; there are many layers of investigation. With each new layer exposed, it’s vital to analyze and inspect the relevant details as moving too quickly with a possible solution could result in breaking KPIs in other areas. 

To assist in the root-cause analysis process, many turn to artificial intelligence and machine learning (AI/ML). However, while excellent and essential, these tools also require deep, hands-on experience and the correct data to generate conclusive solutions — ultimately bringing back up the importance of asking the right questions. 

Finding a Needle in a Needle Stack: How to Work with Data

Aside from asking the right questions, CSPs looking for the root cause of customer experience issues can also get hung up on vast amounts of data. 

Most CSPs have invested significant time, resources and energy into integrating sensors and devices throughout their operations with the aim to collect meaningful data. But instead, the data often has the opposite effect as CSPs are left with massive amounts of information that area struggle to sort through and generate value and actionable insights off of. 

To truly capitalize on their data and the findings it uncovers, CSPs need specialized tools along with industry expertise and advanced domain insight. Fortunately, telco data consultants who have worked with various CSPs in different industries can utilize the best practices and lessons learned from these experiences to truly investigate the large pool of data and uncover the most meaningful insights to determine the correct root cause and best course of action. 

The Customer Really Comes First: How to Start Making Customer-Centric Decisions

The last area where CSPs get stuck is with their decision-making process. Commonly, CSPs focus on a performance-driven approach to decision-making. This puts additional emphasis on areas such as the quality and capabilities of segments of their network. However, this narrow focus on just the radio access network (RAN) or the core can have a net negative impact on the business’s operations when it does not simultaneously better customer experience.

By focusing on the customer end-to-end in the decision-making process, CSPs can ultimately keep customers happy instead of causing their eyes to wander to competitors. But doing so requires a deep understanding and knowledge of the true cause of service complaints to properly address issues affecting the customer experience.

Take the example of customers complaining about dropped calls. At a surface level, it may seem like a solution to invest in optimizing network performance in their area, even if it may risk increasing consumer costs. However, the customer may still be affected by the end-to-end service path, which means that the isolated optimizations might not even improve their experience. Even worse, these increased costs, which may be portrayed as higher service fees as a result of the optimization, could cause the customer to ultimately choose a different provider. 

As a general principle of end-to-end customer-focused decision-making, an investment isn’t effective if it doesn’t better the customer experience. And as a part of that, it’s crucial to uncover and act on true root causes. 

And if issues continue to be unclear or complicated, it may be worthwhile to collaborate with an external partner, who can provide expertise and perspective when investigating root causes. 

Not only would an external provider supply broad industry insight based on ongoing work with various global CSPs, but they also have the domain knowledge and AI/ML expertise needed to ask the right questions, analyze data and uncover the true root cause of any issue – improving end-to-end customer experience by ensuring the customers’ needs are addressed head-on.

Verizon deploys its first Ericsson VRAN cell site

Verizon activated its first Ericsson virtualized cell site (also referred to as Ericsson Cloud RAN), with support from Intel and RedHat providing the processing and cloud-native orchestration functions. 

Ericsson provided its commercial 5G Cloud RAN solution, consisting of a virtualized Central Unit (vCU), a virtualized Distributed Unit (vDU), and radio units. This software-based 5G Cloud RAN solution spans across all of Verizon’s frequency bands, utilizing both FDD and TDD 5G spectrum assets, including Massive MIMO support for C-band and interconnecting previously deployed equipment to enable Ericsson Spectrum Sharing between LTE and NR carriers. The Ericsson Cloud RAN solution offers Verizon efficiency and in network deployment and operational management transformation.

Intel provided its 3rd Gen Intel Xeon Scalable processor, Intel vRAN Dedicated Accelerator ACC100 and Intel Ethernet Network Adapter E810 to deliver the processing, acceleration and connectivity requirements.

Red Hat provided Red Hat OpenShift and Red Hat Advanced Cluster Management for Kubernetes to manage their fleet at scale in collaboration with Red Hat Consulting. As part of the solution, Red Hat Advanced Cluster Management delivers cloud-scale manageability by configuring the 5G RAN using a Distributed Unit (DU) profile, which deploys real-time OS kernel optimizations to run vRAN and other cloud-native application workloads from a single console. Red Hat OpenShift also incorporates zero-touch provisioning to enable distributed deployment at scale required to operate a large scale RAN.

Verizon said Radio Access Network virtualization enables it to rapidly respond to customers’ varied latency and computing needs. 

Virtualizing the far edge of the RAN – the cell sites closest to the customer - is a function of decoupling the hardware and software associated with a cell tower and transitioning the software so that it’s not just stored in the cloud, but is designed for a cloud-native architecture and operation.

Verizon recently confirmed that irecently it has 8,000 cell sites already virtualized in its network with an eye towards virtualizing 20,000 sites.

Verizon begins C-band rollout with Ericsson and Samsung

Verizon confirmed that the installation of C-band equipment from Ericsson and Samsung Electronics Co. is now underway. Ericsson is providing its Antenna-Integrated Radio (AIR) product, which features a massive Multiple Input Multiple Output (MIMO) architecture with Ericsson Uplink Booster and advanced beamforming technology.  Samsung is also supplying Massive MIMO radios and fully virtualized RAN (vRAN) solutions.Verizon secured an...

Verizon completes fully virtualized 5G data session

Verizon recently completed an end-to-end fully virtualized 5G data session in a live network. Verizon's demonstration of virtualization in the Radio Access Network tested technology from several partners, including: Samsung provided its commercial 5G virtualized RAN solution, consisting of a virtualized Central Unit (vCU), a virtualized Distributed Unit (vDU), and radio units. The solution can provide mobile operators with improved efficiency,...

Dell'Oro: Total PON equipment revenue up 27% Y/Y

Total global revenue for the Broadband Access equipment market increased to $4.7 B in 3Q 2022, up 17 percent year-over-year (Y/Y), according to a new report from Dell'Oro Group. Spending on PON equipment continues to fuel the overall market, with revenue for PON OLTs breaking last quarter’s record of $1.3 B in the quarter.

“Thanks to a robust competitive environment and ongoing subsidization efforts, investments in new fiber broadband networks and equipment continue to grow,” said Jeff Heynen, Vice President with Dell’Oro Group. “It will be interesting to see if this momentum and resilience will carry into the fourth quarter and 2023, given the macroeconomic headwinds,” explained Heynen.

Additional highlights from the 3Q 2022 Broadband Access and Home Networking quarterly report:

  • Total XGS-PON ONT unit shipments exceeded 1 M for the second straight quarter, reaching 1.5 M worldwide.
  • Total cable access concentrator revenue was flat Y/Y at $259 M. Remote PHY devices and remote OLTs both reached record levels in the quarter, as cable operators continue to expand their DAA and fiber initiatives.
  • Total cable CPE units increased 5 percent Y/Y, as vendors were once again able to fulfill orders that had been backlogged for multiple quarters. Total revenue was up 9 percent Y/Y as higher-end DOCSIS 3.1 gateways are the percentage of total unit shipments.

Dell'Oro: Double-Digit growth in N America’s Service Provider router market

Growth in North America and the Caribbean and Latin America (CALA) regions continued accelerating in 3Q 2022 leading the worldwide Service Provider (SP) Router and Aggregation Switch market 5 percent year-over-year growth, according to Dell'Oro Group.

“North America’s continued upward trend in 3Q22 was driven by SPs upgrading networks to gain the advantages of 400 Gbps technologies and expanding their networks to accommodate growing traffic. Demand for network capacity continues to increase, driven by 5G, IoT, pervasive video, and other technology trends,” said Ivaylo Peev, Senior Analyst at Dell’Oro Group. “From a worldwide perspective, solid revenue growth in the combined SP Edge Router and Aggregation Switch segment was partially offset by flat performance in the SP Core Router segment. The fast-growing adoption of Edge Router and line cards based on the newest ASICs, which support 400 Gbps connections and offer the highest available bandwidth, was the key reason for the Edge Router segment growth in 3Q 2022,” added Peev.

Additional highlights from Dell’Oro Group’s 3Q 2022 Service Provider Router and Switch Report:

  • High (double-digit) revenue growth in North America and CALA was diminished by declines in Europe, Middle East and Africa (EMEA), and Asia Pacific (APAC).
  • Market growth was strong in the SP Edge Router and Aggregation Switch segment, while the Core Router segment trended flat.
  • Supply chain constraints eased somewhat in the quarter, improving access to components and enabling SP Router vendors to work through some of their order backlogs.
  • Among the top five SP Router Suppliers, Cisco and Juniper recorded solid revenue growth, while Huawei and ZTE declined. Nokia had a strong quarter based on revenue reported in Euros (EUR), but the adverse exchange rate from EUR to U.S. dollars (USD) resulted in a revenue decline in USD.

Marvell launches secure 1.6T Ethernet PHY

Marvell Technology has begun sampling a 5nm 1.6T Ethernet PHY with 100G I/O capability, featuring built-in Media Access Control security (MACsec) and full support for Precision Time Protocol (PTP).

MACsec allows data center operators to enable hardware-based link-layer security and PTP enables Ethernet networks to support ultra-reliable timing- critical services. The X9340P and the previously announced Alaska C X93160 PHY together provide a pin-compatible, flexible platform for retiming, gearboxing, encryption and timing applications for speeds up to 800GbE.

Marvell's new "Alaska" X9340P is built on the company's PAM4-based 112 Gbps SerDes technology and delivers 40% lower power-per-bit compared to PHYs using 56Gbps PAM4-based SerDes technology. The Alaska C X9340P PHY incorporates the IEEE 802.1AE 256-bit MACsec capability in the PHY, providing for flexible encryption deployment while eliminating the cost and power burden of including this functionality in the switch ASIC. Marvell is supporting the Switch Abstraction Interface (SAI) driven by the Open Compute Project (OCP) on both the X9340P and X93160 devices to enable the vendor-independent control of networking equipment elements demanded by leading cloud data center customers.  

The new PHY also integrates support for PTP for timing-critical use cases in 5G and telecom applications. The stringent timing requirements of 5G networks and applications served by them are driving the timing accuracy that needs to be delivered by networks supporting these services. The inclusion of Class C PTP timestamping enables the support of sub-microsecond end-to-end clock synchronization accuracy needed for 5G Ultra-Reliable Low-Latency Communications (URLLC) services that enable mission-critical use cases such as industrial automation.

“100G PAM4 I/Os and 800 Gigabit Ethernet are the next leap forward for the industry,” said Venu Balasubramonian, vice president of product marketing, High Speed Connectivity and PHY Business Unit at Marvell. “Our newest 1.6T Alaska C PHY solidifies Marvell’s position as the silicon provider with one of the broadest and most advanced high-speed Ethernet PHY portfolios for cloud data center and carrier markets.

Marvell intros 400G/800G PAM4 DSPs for Active Electrical Cables

Marvell introduced its Alaska A PAM4 DSP family for Active Electrical Cables (AECs) designed for data center interconnects by hyperscale customers.The new Alaska 400G/800G DSPs, fabricated in 6nm process, leverages Marvell’s PAM4 DSP technology to improve signal integrity in copper, short reach AECs to address emerging 100G/lane adoption in cloud data center interconnect architectures. The new Alaska 400G/800G DSPs add to Marvell’s comprehensive...

Marvell's Matt Murphy appoint chair of Global Semiconductor Alliance

 The Global Semiconductor Alliance (GSA) Board of Directors appointed Matt Murphy as its new chair.

Matt Murphy is President and Chief Executive Officer of Marvell Technology. He has led the company since joining in July 2016 and also serves as a member of its Board of Directors. Prior to joining Marvell, Matt worked for Maxim Integrated, where he advanced through a series of business leadership roles over two decades. Most recently, he served as Executive Vice President of Business Units and Sales & Marketing, overseeing all product development and go-to-market activities.

“I am honored to become Chair of GSA’s Board of Directors and look forward to partnering with my fellow Board members and other partners to further advance the global semiconductor ecosystem,” said Matt Murphy. “Semiconductors have proven to be an indispensable part of the global economy and there has never been a more important time for the entire ecosystem to come together and collaborate on a global level. The semiconductor industry continues to be a catalyst for technological progress and the world is counting on this group to fuel innovation like never before.”