Improving SOC Insight into Customer Experience with Accedian Skylight

Available Languages

Download Options

  • PDF
    (1.3 MB)
    View with Adobe Reader on a variety of devices
Updated:August 30, 2021

Bias-Free Language

The documentation set for this product strives to use bias-free language. For the purposes of this documentation set, bias-free is defined as language that does not imply discrimination based on age, disability, gender, racial identity, ethnic identity, sexual orientation, socioeconomic status, and intersectionality. Exceptions may be present in the documentation due to language that is hardcoded in the user interfaces of the product software, language used based on RFP documentation, or language that is used by a referenced third-party product. Learn more about how Cisco is using Inclusive Language.

Available Languages

Download Options

  • PDF
    (1.3 MB)
    View with Adobe Reader on a variety of devices
Updated:August 30, 2021
 

 

1. Executive Summary

Communications Service Providers (CSPs), driven by organizational challenges, new technology, and service quality demands, are evolving to become customer-centric businesses. New technology such as 5G is designed to meet the new customer requirements for guarenteed performance and security. This has increased focus on how networking, mobility, and applications relate to the end customer’s Quality of Experience (QoE) and satisfaction, a key driver in generating new revenues and instilling customer loyalty.

Skylight provides unified application and network performance management

The growing pressure to deliver an excellent customer experience, reduce customer churn, and develop new revenue sources means CSPs are transforming legacy processes and systems into an architecture that can proactively manage actual QoE in real time. This has led to a shift in performance management focus from reactive to proactive, real-time data streaming and analytics where the ability to predictively act on QoE issues is key.

This paper examines how Service Operation Centers (SOCs) can support that proactive approach to managing customer experience while adopting new technology such as 5G and the potential challenges.

Important considerations include the timeliness and quality of the Key Performance Indicator (KPI) and Key Quality Indicator (KQI) data used to calculate the Customer Experience Index, which is indicative of a customer’s “happiness with a particular service.”

The main challenges of deploying an SOC lie in resource costs (financial or other), how long it will take to implement and integrate the SOC into existing systems and architecture, and the need to transform business processes.

Finally, we discuss how Accedian Skylight can help with the CEI analysis process across the networking and application layers of the overall architecture.

Accedian’s virtual instrumentation supports dynamic performance management measurement capabilities, such  as IoT or business-critical services in a 5G network slice.

With 5G networks and services, the quality of performance data will not only depend on the granularity,  precision, and accuracy of the performance management system, but also on the ability to embed performance  management software to dynamically measure network slices and service chains on demand “per customer.”

Analytics are needed to process that streaming data and feed service quality insight in real time into  orchestration and automation systems.

Skylight analytics supports closed-loop automation, machine learning, actionable insight through service  performance-monitoring visualization, and data-cleaning capabilities using industry-based standards.

We examine how Accedian Skylight fits with the SOC model by providing passive and active monitoring, analytics, and orchestration of service quality management in the network and application domain.

The paper also discusses why Skylight is a good fit to support existing SOC deployments or to help service providers migrate to an SOC—without significant resource costs.

2. Synopsis

An industry blueprint or gold standard for designing the “ideal” SOC architecture is “a work in progress.” Instead, the lead vendor typically implements what’s available from their portfolio, leading to varied deployment models and success rates.

Service providers therefore must consider their big picture strategy and take a vendor-agnostic approach to define what to include in an overall SOC architecture.

It’s relatively straightforward to specify SOC components and how they relate to each other. It is much more complex to define mechanisms, analytics, data, and how the architecture will work together to calculate a “realistic” Customer Experience Index (CEI). (See page 8 for more discussion.)

A key ingredient is the quality of performance management and analytics data ingested by the system to create actionable insight related to the customer experience. (See Figure 1.)

Operational & planning lifecycle operations migrate towards service-based model

Figure 1.            

Operational and planning lifecycle operations migrate towards service-based model

Essentially a SOC is designed to provide the customer’s view of Quality of Experience (QoE) as the central point of reference. Key components impacting customer QoE and satisfaction are network and service layer availability and reliability, which can be determined with accurate performance management data.

If an application is slow or a service is down and the network is not performing well, customers will be unhappy, independent of the other components such as billing and customer care. Often the network and service portion of the calculation for CEI can have a greater impact on the overall state of “happiness of the customer.”

Network reliability and service availability have significant impact on customer experience

Figure 2.            

Network reliability and service availability have significant impact on customer experience

Accedian Skylight enhances the capabilities of a SOC by providing an accurate view of customer QoE based on high-quality performance data from the network, application, and services layers. Skylight uses active and passive monitoring, analytics, and machine learning to feed timely and accurate KPI and KQI data specified in a SOC model. Skylight also provides visualization and reporting capabilities based on business rules and APIs to other systems and supports closed-loop automation.

Digital IT processes and end-to-end performance management

Digital transformation of IT processes and functions that are being redesigned to be digital end-to-end also have to be considered as part of a service provider’s SOC strategy. Top-to-bottom application and network visibility, at each layer, will be a key factor in understanding any impact on the customer experience.

From a performance management perspective this means that IT and network operations will require a full-stack view. This will provide control, actionable insight and visibility, and the right performance metrics to ensure that the end-to-end network, service, and application lifecycle are working correctly. This also means that IT professionals will need to become network savvy and network professionals will need to understand the needs of IT teams.

This visibility can be achieved by converging active and passive monitoring to provide a big-picture view, showing all components involved in end-to-end performance management.

Skylight active and passive monitoring provide actionable insight into how customers are experiencing the overall service performance (combined network and application QoE) across IT infrastructure, applications, and network infrastructure.

When there is a QoE degradation, the solution pinpoints in seconds whether the issue is due to the application code, the IT infrastructure, or the network. Having a single lens view of the combined application and network performance provides a unified understanding of end-to-end quality from the customer perspective. This improved insight enables faster problem resolution compared to tools that simply look at application and network performance in silos.

Skylight provides unified application and network performance management

Figure 3.            

Skylight provides unified application and network performance management

The main value of Accedian Skylight lies in being able to provide the “reality” of what’s happening at the service and network layers to the SOC as opposed to the perception that customers are having a good quality experience and Service-Level Agreements (SLAs) are being met. This allows for more realistic customer experience KPIs and KQIs to flow into the SOC engine. By providing preanalyzed data and machine learning capabilities through Skylight analytics, the information shared with the SOC and orchestrators is further enhanced. Clean performance data that correlates exactly to the business rules and network services being used by customers is the key to successfully calculating a realistic QEI based on the network and services layer impacts.

High-quality and timely data is extremely important in a SOC to get an accurate view of service and network quality from the customers’ viewpoint. Performance management tools on the market today struggle to handle the new requirements driven by customer demand for better quality and more personalized services, which can impact an operator’s ability to compete effectively.

Thinking customers are happy vs. knowing is the difference between the perception and the reality of your visibility and thus your control of your network. The focus of a SOC and related performance management systems should be on showing what “is really” going on in the network and service layer as opposed to showing only what is going on in the network.

In summary, the quality of the information and performance data ingested by the SOC will determine how accurately the real experience of a customer is measured. Even the smallest of fluctuations in the network can have significant impact on the service and network quality and customer experience.

This paper discusses SOC architecture and how Accedian Skylight provides high-quality analytics and performance data, in order to help improve customer satisfaction and drive sustainable revenues for service providers.

3. Network Management Evolution

3.1 Drivers for change

Communications Service Provider (CSP) networks are key enablers of digital service delivery. To stay relevant, CSPs have to be proactive in providing the high-quality services their customers expect.

There is growing pressure to deliver an excellent customer experience, reduce customer churn, and develop new revenue sources. Customers will only buy more if they are happy with the service. For CSPs, this means transforming legacy processes and systems to measure and proactively manage the actual quality of services customers experience in real time. Doing so involves shifting the focus from reacting to network issues to being able to predict what could happen; the goal is being able to avoid subscriber impact by taking proactive action.

Traditional Network Operations Center (NOC) tools and processes are highly network-centric and so lack service, customer, and business context. (How is the service experience through the eyes of customers?) When customers call in saying, “I’m trying to watch TV on my iPad and can’t because it keeps lagging, I’m not happy with the service,” CSPs need to be able to see the customer’s “service layer experience.” In order to understand why the video is downloading too slowly or not at all, it’s necessary to go beyond merely monitoring resource and network layers.

3.2 The rise of the SOC

Network teams are more focused than ever on Service Quality Management (SQM), which involves implementing a SOC and new analytics tools capable of measuring and proactively managing service quality.

Overview of SOC in organization structure

Figure 4.            

Overview of SOC in organization structure

These SOCs are designed to bridge the worlds of network operations, commercial teams, and automation capabilities.

In the future, SOCs will play even bigger roles in identifying new service and upsell opportunities and proactively engaging customers based on their particular interests and behaviors.

SOCs are focused on monitoring, in real time, all the service components for each customer. CSPs can then use this insight to prioritize and resolve issues that may impact high-value VIP customers or customers that may be at risk of churn. This requires having a single view in real time of the subscriber experience, across all networks (edge and core, fixed and mobile), devices and services, applications, and related domains in order to understand the experience of the end user.

3.3 Customer Experience Management plays a key role in transformation

Reducing customer churn and retaining customers is a key driver for better understanding the customer experience. That’s hardly surprising, as it is significantly more expensive to win a new customer than to maintain an existing one. Churn reduction is closely related to the need to improve customer satisfaction and Net Promoter Score (NPS). Top-level managers within operators are now paying close attention to customer satisfaction and NPS targets, including ratings for each country’s operations.

From network-centric to customer-centric operations

Figure 5.            

From network-centric to customer-centric operations

Leading CSPs are increasingly deploying a Customer Experience Index (CEI) correlated with NPS to understand the current state of customer experience and satisfaction combined. The CEI is used in SOCs to determine service quality degradations or understand the types of customers impacted by an outage.

CEM solutions are becoming more layered and based on big data architectures designed to store, analyze, and correlate all relevant customer data to get an end- to-end view from Customer Relationship Management (CRM), Operations Support Systems (OSSs), and Business Support Systems (BSSs).

All these developments ensure that network performance management is based on algorithms that provide high-quality performance data and metrics correlated to identify issues that impact the customer experience. What matters most is the quality of the data ingested and how it is analyzed.

3.4 Reliable network and service data is the foundation of SOC CEM systems

The most important mission of a SOC is to understand how customers are experiencing services. Getting a true picture of that is based on the quality of data ingested by the SOC and the capabilities of its analytics engine. One key component needed is the ability to collect, ingest, and visualize performance management data from Layer 2 to Layer 7. Without this level of detail, we’re missing the truth about the customer “experience.”

In the past, looking at network metrics was enough to resolve many of the issues that customers were experiencing. But today the granularity, accuracy, and precision of data—and turning that into actionable insight—is the key to any successful SOC.

From SOC to automation

Figure 6.            

From SOC to automation

A “successful” SOC, therefore, uses a single CEI factoring in business, operational, and network processes, with network QoE assigned an appropriate weight. The key here is being able to measure the end-to-end journey of the “service” and how each component impacts the overall customer experience or CEI.

All of these developments considered, network, service, and application performance management becomes a science rather than an art, using algorithms to predict and control service quality proactively—and by proxy, the user experience—rather than compensating for or fixing user experience failures after they have occurred.

Operators are prioritizing network and service quality improvements within their CEM programs because, as one European incumbent operator puts it, “the network better be top-notch and highly reliable, or nothing else matters.” A single CEI that drives network optimization is highly valuable, but many operators find it difficult to provide common visibility and unify efforts across network and service delivery organizations.

3.5 Predictive, real-time analytics play a key role in CEM automation

Automated service quality management is the ultimate goal. Employees sitting behind desks in the SOC looking at only Key Quality Indicators (KQIs) and CEI will not be enough to manage dynamic, on-demand networks and services in real time.

With software-defined programmable networks starting to go live, CSPs realize that operations teams will require highly accurate and trusted analytics to control service quality and automate policy on service orchestration. Predictive analytics and machine learning will also be deployed to identify service anomalies or service-impacting events and allow such events to be proactively managed.

3.6 SOC transformation

Communications Service Providers (CSPs) are setting up Service Operation Centers (SOCs) to monitor all the services it provides to customers in real time, end-to-end, at the service layer. They are using a Customer Experience Index (CEI), correlated with Net Promoter Score (NPS), to manage customer segments (e.g., VIP customers). This allows CSPs to assess service degradation by cell site and segment groups of impacted customers based on ARPU and how “valuable” they are.

This strategy is changing business processes across network operations, customer care, and planning teams. For example, they can plan upgrades based on improving quality of experience and not just technical parameters. It also enables them to reduce calls to customer care by prioritizing fixes for network problems that have the greatest impact on customers. A customer with a digital self-care app or portal can then see if there is a problem and that the service provider is working on it. And when a customer calls customer care, the CSP knows about the customer’s problem and root cause.

CSPs are also using big data analytics and machine learning to anticipate incidents and identify proactive actions to improve the customer experience with the ultimate goal to provide excellent network and service performance.

“We are setting up service operating centers where we will monitor, in real time, all the services we provide to customers. Every operator needs this now— with mobile video and voice services becoming more complex. If the network is green, and the customer has a problem; we have problems with current tools. What we need is real-time customer vision, because only monitoring the resource and network layers is not enough.”

CTO of tier 1 European operator

4. Challenges in SOC Transformation

The transition from NOC to SOC changes the focus from ensuring the smooth operation of the network to understanding and safeguarding the quality experienced in key consumer services and by individual high-revenue customers.

The usual stumbling block is the lack of reliable, integrated data. Fundamentally, a SOC can’t function without accurate and up-to-date knowledge of the changing dependencies and state between a customer, their services, and the entire end-to-end network.

This data usually exists, but it is spread out across a multitude of BSS, OSS, and network systems, and a key challenge in transitioning to a SOC is to quickly bring this data together into a single, coherent, real-time view.

Moving to a customer-centric SOC approach requires a huge collaborative effort, not only on the technical architecture, as it also requires the entire organization to develop a more customer-centric culture and new integrated organizational process models.

4.1 What makes SOC implementation challenging

      Cost: implementation, operation, upgrades, data management, storage, and resources.

      Organizational change and business processes: what’s needed to support the CEM and SOC model.

      Customer and network, service, and application performance architecture creation: a large and complex endeavor.

      Data: what’s collected must be clean, accurate, timely, and meaningful.

      Human resources: lack of highly skilled employees (who are able to make sense of data) with experience in data science, machine learning, and analytics.

      Vendor lock-in: addressing this possibility once a particular vendor’s CEM system, including algorithms, is embedded into the SOC.

      KPIs: how will these indicators be collected and how will service quality management be deployed to address the dynamic nature of NFV, SDN, cloud networks, and new digital services.

The question then becomes: “Where can I start this transformation process with minimal impact on cost? How can I start reaping the benefits ASAP? What can I do on ‘Day One’ to make a real difference?” The answer lies in being able to accurately measure and correlate the entire end-to-end service network path with high-quality performance management and analytics data.

4.2 Importance of timely, relevant, accurate KPIs and KQIs

Understanding the challenges is important, but so too is understanding what type of architecture a service provider needs to support:

      Creation and management of scheduled/ performance-based service segments

      Multilayer, multislice (5G) management for per- service QoE

      Continuous monitoring and reporting of service quality and SLA parameters

      A closed-loop architecture that orchestrates network slices in real time

      Keeping track of the behavior of user and applications

      Future-proff for 5g and other technologies that will be customer centric

      Correlation of all performance metrics—cross- functional and multilayer

These requirements and capabilities point to the need for context-based service assurance and Service Quality Management (SQM) architecture that can provide a SOC with:

      Network, service, and application performance

      Real-time network and service monitoring

      Real-time application monitoring and analytics

      Multilayer troubleshooting

      Dynamic and automated network topology awareness for performance management

      Cross-domain visibility and correlation

      Capacity optimization and management

      Support for automated service delivery

      Real-time root-cause analysis

It’s clear that the overall value of a SOC/CEM model is based on the visibility and quality of insight that network, application, and service performance management tools provide.

“We believe that we need to have tools that can correlate performance layer by layer; does a problem originate at Layer ZERO or further up the stack? We are actively looking into solutions like that. Our main goal is to visualize the service layer and network layers. Our tools now are not correlated; we have to manage network management systems (NMSs) from multiple vendors. If we need to troubleshoot, we have to manually do that. If we could get a more unified common network management layer, that would really help.

We need end-to-end monitoring of our networks (wireless and wireline) to deliver end-to-end SLAs—visibility into all network layers, including transportation, microwave, and other network segments. This is important.”

Tier 1 Asia/Pacific national service provider

5. Digital Transformation and the SOC

Performance management in the SOC is more than just the network. As CSPs transform IT systems and processes to be digital end to end and move key IT applications to the cloud, there is a need to actively monitor the connectivity, but also to have a full-stack view of application visibility and performance. This measurement data can then be correlated to comprehensively assure end-to-end service performance.

Performance management and ensuring that the service and customer QoE is sustained throughout the digital transformation process is a key challenge that requires a unified view of a customer’s journey across the IT, network, application, and service domains.

This also means that the components (CRM, OSS, BSS, etc.) feeding data into the SOC are part of the digital transformation process, and also that SLAs for these components must be monitored.

Application Performance Management (APM) and Network Performance Management (NPM) together with a granular, carrier-grade network path and transport layer measurement architecture are needed. The performance data collected from each component will build the analytics data to generate the “real” actionable insight needed to align with business goals.

Active monitoring that can identify network layer- related issues at a very precise and granular level is provided by Skylight. This combined APM and NPM insight provides complete visibility from Layer 2 to Layer 7.

Through passive monitoring, Skylight can visualize all of the transactional and relational connections of applications. In fact, Skylight is designed to monitor how the end user is experiencing the application performance (application QoE) as it traverses the IT infrastructure and the network.

When there is a QoE degradation, Skylight will pinpoint in seconds whether the issue is due to the application code, the IT infrastructure, or the network.

Skylight’s ability to present a single pane of glass on performance differentiates it from tools that simply look at the performance of the network or applications in silos.

QoE cannot be realistically measured without understanding precisely how the service and applications are behaving in relation to how the customer is actually experiencing the service. In short, we need full-stack visibility from the network layer right up to and including the service and application layer.

Just as IT and network skills are converging, application and network performance is converging in order to understand and assure the customer experience. Without truly seeing how each of them impact one another and the actual customer experience, increased operational costs and internal resources are then needed to solve problems as a result. Revenue loss from customer churn is also likely if customers have problems accessing their applications even though the network is working for other purposes.

6. Accedian Skylight

6.1 Skylight for customer experience and SOC management

Skylight is designed to provide a customer-centric view of the network and application performance. It can be integrated into SOC platforms to provide an even more complete view of the network and the customer experience. For an existing CEM/ SOC deployment, Skylight is a low-cost, high-value investment to improve customer experience.

6.2 Network and application analytics correlated with metadata for actionable insight

Skylight performance analytics feeds events and data (raw or aggregated/customer correlated) into the SOC platform to provide a more complete view of the customer experience. The SOC can then take this processed data to either perform closed-loop actions directly on the network or can use the Skylight platform to increase the level of testing/granularity in order to gain an even more precise view of performance and the customer experience.

6.3 Skylight analytics

In addition, Skylight can integrate into a customer portal (using Single Sign-On [SSO] into the SOC user interface) to provide customers a direct view of quality of experience.

Other tools in the network, service, and application performance and monitoring space provide the ability to graph and report on data collected from the network but fall short on providing real insight into this data.

Skylight analytics was built from the ground up with one mission: to provide actionable insight about how the customer is experiencing their service, all the way from the network and services layer to application layer.

For data ingress, Skylight analytics offers an open data pipeline for integration of any time series data from the network. All Skylight performance management capabilities are supported out of the box.

Skylight analytics can also support SNMP, CSV/XML/JSON files, streaming telemetry data using gRPC, and other similar data formats and protocols. This means, for example, that performance management data can be enriched with other sources of data and correlated to enable faster troubleshooting and root-cause analysis.

Once data is stored, it can be easily enriched. Either through the user interface, or with RESTful APIs, the system is able to tag each object with a “domain” to control visibility and access to the object. Each object can also be enriched with metadata. Such metadata can be leveraged in queries, dashboards, and reports in Skylight analytics to provide additional slices and views of the network and to provide additional power for correlation and prediction algorithms.

For data egress to other systems, Skylight analytics supports an open data pipeline with JSON over WebSockets and gRPC, allowing access for both raw and synthetic/ generated data (alarms, events, etc.) from Skylight into other systems.

Finally, Skylight analytics supports RESTFul APIs for integrating other application functions. Any actions that can be performed with the user interface (e.g., configuration, adding domains or metadata, and so on) are also available from an open and documented REST/JSON/API. This allows full integration with the BSS/OSS layer and opens any function in Skylight analytics to scripting.

6.4 Skylight overview

Skylight is a fully virtualized (software-based) performance assurance platform that delivers network-wide visibility. The software component can be used with Skylight’s FPGA-based, hardware-assist modules where it’s appropriate or beneficial to retain and improve the precision, granularity, and breadth of information traditionally only available with hardware-centric solutions. Skylight components include:

      Performance sensors (control): This virtualized active testing and monitoring sessions generator can be deployed in minutes to monitor service performance. Performance modules are managed and controlled by Skylight sensors, which provide the NFV compute and control for active test sessions. Skylight sensors can also act alone without modules to actuate and reflect active test sessions in a software-only session.

      Performance sensors (on-premises and cloud): This adds passive capture at Layers 4 to 7 to the Skylight portfolio. This can parse all network traffic but is most efficiently used to only store and transmit key metadata to a central repository—thus providing a compact, scalable, and fully virtualized method of capturing the user experience at the application layer.

      Skylight orchestrator: Orchestrator controls the entire Skylight portfolio, either using its own user interface or RESTful APIs. It is easy to set up and automate new tests and monitor existing tests using Skylight orchestrator.

      Performance analytics: Skylight analytics collects data from the entire Skylight platform (performance sensors [active monitoring and passive capture] and software tests) and combines this with other sources of network data to provide a complete view of the customer experience.

Combined, these components provide the flexibility and interoperability operators need to measure QoE, integrated with existing and future network environments. Performance sensors, modules, SFP compute sensors, and performance elements (network interface devices) can be combined to provide complete, active test coverage of the mobile network from core to aggregation to backhaul and all points.

Skylight’s simplified, virtual architecture means it can be deployed quickly and transparently to provide an instant measure of the performance and quality of experience of the end-to-end service. Each component provides the SOC with precise, granular, real-time, and multilayer data. The data can then be correlated with RAN, CRM, and other data.

Analytics acts as the bridge between Skylight and the SOC to provide the first level of preanalytics data processing, visualization, and machine learning capabilities for closed-loop service quality management automation.

Skylight solution overview

Figure 7.            

Skylight solution overview

7. How Skylight Can Improve your SOC

7.1 Skylight capabilities

Skylight provides high-quality performance metrics to support the overall architecture of a SOC. It can work with, and integrate into, any SOC, NOC, or CEM architecture—for example, Amdocs, Huawei, Ericsson, or Nokia—that uses standards-based interfaces and protocols for network measurement.

Even without deploying a heavyweight SOC CEM system, Skylight performance analytics provides visualization, reporting, and integration capabilities. This helps operators identify key QoE issues by correlating them with the performance management data from Skylight and third-party systems.

Accedian’s Skylight addresses four main areas:

      Passive monitoring

      Active monitoring

      Performance analytics and machine learning

      Orchestration and closed-loop automation

The solution’s inherent design works well in an NFV/ SDN virtualized environment with or without physical instrumentation—depending on the precision and accuracy of the data required for a particular service segment.

The overall architecture works this way:

      Performance sensors and modules provide application and network performance management data all the way from Layer 2 to Layer 7.

      Skylight orchestrator acts as the central point for active test deployments and can also automate performance-monitoring instantiation via RESTful APIs.

      Skylight orchestrator manages performance data and monitoring elements throughout the network.

      Skylight orchestrator manages inter-Skylight performance-monitoring instantiation.

      Skylight provides application-layer metrics, correlated in Skylight analytics to give the big picture impact on customers QoE: how services are behaving in relation to network metrics and the service path.

Skylight performance analytics solution

Figure 8.            

Skylight performance analytics solution

7.2 Quadrants covered by Accedian Skylight

Active monitoring

      Skylight orchestrator is a centralized tool used to discover, configure, and receive active performance-monitoring data. A single API can be used for all active monitoring, via a mediation/abstraction layer to the devices and software in the network.

      Skylight control sensor is a virtual (software), standards- based performance-monitoring session generator that can be deployed in minutes to enable assurance monitoring at a granularity aligned with the service being delivered.

Analytics and machine learning

      Data such as TCP efficiency is used to glean detailed user experience info so the network can be managed effectively.

      Accurate, real-time performance management data is sufficiently “clean” for smart analytics and machine learning. The Skylight flowmeter feature provides the real-time bandwidth utilization required for closed-loop applications (for example, triggering a packet capture based on a burst to determine the cause of the burst).

      Skylight application data is accurate and real time—a key requirement for smart analytics and machine learning.

Passive monitoring

      Skylight orchestrator helps to collect detailed user experience information for monitoring and adjusting the network to optimize QoE from an active perspective.

      Skylight consists of a passive monitoring tool that complements Skylight active monitoring. This virtual, lightweight Layer 3-7 solution captures north-south and east-west traffic and processes that data in real time, so only 0.2% to 0.5% of traffic is needed to provide complete and detailed (transactional) application information metadata.

      The Skylight flow broker feature remotely captures packets to provide valuable application and network (TCP) statistics derived from actual customer traffic.

Orchestration and automation

      Skylight orchestrator acts as a real-time central touch point for north and southbound communications; this is a crucial part of automaton and closed-loop applications.

      The Skylight control sensor module executes real-time automation layer requests.

      Skylight complements network, service, and application performance monitoring with real-time feedback to validate the user experience after automation changes have been executed.

7.3 Benefits of integrating the Accedian Skylight QoE approach for SOC architecture

It is clear that integrated SOC, NOC, and CEM systems will play a key role in ensuring service providers can deliver high QoE on services by understanding the behavior, experience, and expectations of their customers, and differentiate from competitors.

How Skylight fits into that:

      While it’s not a fully-fledged CEM solution, Skylight provides the critical quality performance management data that CEM or SOC systems need to define, evaluate, and manage a customer’s experience.

      The solution works with and integrates into any SOC CEM system to provide a high-quality, multivendor view of the network, service, and application layers. It effectively provides good, clean, usable data on the networking, service, and application component into the SOC.

      Skylight provides a vendor-agnostic way to collect performance management data, identify customer QoE, and gain a nearly real-time view of how customers are experiencing services.

      Whether issues originate in the backhaul network, RAN connectivity, the application layer, or elsewhere, Skylight’s unified architecture and features provide the necessary level of detail to ensure end-to-end Service-Level Agreements (SLAs) are being met.

      Key components of Skylight can feed and integrate into an orchestration and virtualization environment to provide necessary KPI and KQI triggers to support closed-loop automation, thereby ensuring service availability and quality.

      In addition to providing high-quality performance management data and insight, Skylight employs a lightweight implementation model that supports NFV Infrastructure (NFVI) and can be configured and instantiated in a much shorter time period than the underlying probe architecture that a SOC or CEM typically relies on.

      Skylight ensures that a SOC can measure the true customer experience in the context of network, service, and application performance. It does this by providing data that is representative of customer traffic 24x7 and measures diverse aspects of a service, ranging from service activation testing to monitoring Layer 3-7 application behavior.

7.4 Value of Accedian Skylight solution vs. heavyweight SOC CEM systems

      Skylight provides a fully integrated, easy-to-implement solution across active testing and monitoring, passive monitoring, analytics and machine learning, and orchestration. Typically, this would require integration work and interoperability between different systems and a mediation layer to access data from existing tools and systems.

      Currently most SOC models are based on vendor-specific solutions. In order to achieve a timely and successful deployment, an operator must select a vendor that can integrate the full architecture spectrum—including the building, possible operation, and management of the entire process to move from an NOC to an SOC.

      Deploying a fully-fledged SOC means opening up the CSP’s entire “DNA” to the vendor, for such matters as supporting or managing OSS rationalization and NOC optimization.

      Accedian Skylight is very easy and fast to deploy compared to a basic CEM or SOC solution. In one case, for example, it took only 90 days to deploy 180,000 LTE measurement sessions to measure an operator network with 10,000 eNodesBs

      On average, a comprehensive SOC architecture once deployed can reduce OpEx up to 20% but the initial CapEx required is high compared to Accedian’s approach.

      Heavyweight SOC CEM systems typically have a very large footprint and require significant hardware deployment throughout the network to collect and analyze data.

      In short, Skylight can complement and/or replace components of the SOC in the application, service, and network performance areas depending on the requirements of the overall architecture.

8. The Cisco and Accedian SolutionsPlus Partner Advantage

Accedian is an official member of the Cisco® DevOps SolutionsPlus Program. This means that CSPs worldwide can approach Cisco and approved Cisco partners to purchase Accedian Skylight as a validated solution and addition to Cisco Crosswork.

Cisco provides comprehensive lifecycle service automation to enable faster design and delivery of high-quality customer services and experiences, accelerating digital transformation for hybrid networks.

Cisco helps automate processes by bringing together closed-loop visibility, insights, and actions to manage business resiliency and IT agility. Cisco collaborates with partners like Accedian with network service quality and real user experience visibility, insights, and actionable events to inform and enrich the automation of the network to deliver digital business outcomes and quality customer experiences.

9. Conclusion

Accedian Skylight fits into the SOC model based on its ability to provide high-quality, reliable performance management data where it is needed in a timely and lightweight fashion.

How does Accedian address end- to-end service management based on quality of experience from a user perspective?

      Active monitoring – Lightweight, virtual active test agents can be deployed as part of service activation testing to assure performance using synthetic traffic. This could be deployed directly to Customer Premises Equipment (CPE) to provide an end-to-end view.

      Passive monitoring – What Accedian Skylight does extremely well is monitor how the end user is experiencing the application performance (application QoE assurance) as it traverses the IT infrastructure and the network. When there is a QoE degradation, Skylight will pinpoint in seconds whether the issue is due to the application code, the IT infrastructure, or the network. Skylight’s ability to provide a single-lens view of performance differentiates it from tools that simply look at the performance of each in isolation.

      Analytics and machine learning – Skylight performance management data is accurate, multilayer, real time, and “clean” to support analytics, machine learning, and closed-loop automation. Skylight analytics can correlate multiple data sources, attach metadata tags, and even decide to capture additional packets in order to speed up time to insight and problem resolution.

How does Accedian support current and future needs for analytics and metadata correlation?

Skylight analytics can generate and feed actionable insight into the SOC platform to provide a more complete view of the customer experience. The SOC can then take this insight to either perform closed-loop actions directly on the network or instruct Skylight to increase the level of testing or granularity to get an even more precise view of performance and the customer experience.

Once data is stored in Skylight analytics, it can be easily enriched. Through the user interface or through RESTful APIs, the system is able tag each object with a “domain” to control visibility and access to the object. Each object can also be enriched with metadata. Such metadata can be leveraged in queries, dashboards, and reports to provide additional network views and insight.

For data egress to other systems, Skylight analytics supports an open data pipeline with JSON over WebSockets and gRPC, allowing other systems to access both raw and synthetic/generated data (alarms, events, etc.).

Finally, Skylight analytics supports RESTFul APIs for integrating other application functions. Any actions that can be performed with the user interface (e.g., configuration, adding domains or metadata, and so on) are also available from an open and documented REST/ JSON API. This allows full integration with the BSS/ OSS layer and opens any function in

Skylight analytics to scripting. Moreover, correlation can be determined by the system across all network and application domains, and additionally, all data elements can be ingested.

How can Accedian Skylight help to create new sources of revenue (managed IT service) or improve the customer experience using performance monitoring and data insight?

Accedian Skylight provides highly accurate, real-time network performance management and application performance insight on enterprise cloud services from data center to the end user. This makes it possible to target new revenue opportunities such as SD-WAN, enhanced application reporting and performance management as a service, and the next wave of universal CPE-based services (virtual firewalls, unified communications, WAN optimization, etc.). Skylight can also be integrated with other data sources and presented in online customer portals to give customers a direct 24x7 view of application, service, and network performance. Customer satisfaction with service performance, faster problem resolution, and ease of use of the customer portal will encourage enterprises to buy additional managed services from operators.

How will Accedian cope with a rapidly evolving network that grows exponentially in number of network elements (driven by IoT, 5G, NFV, etc.)?

Accedian’s virtual instrumentation supports dynamic performance management measurement capabilities, such as IoT or business-critical services in a 5G network slice.

Skylight provides very granular, precise, accurate data that can be dynamically adjusted though RESTful API integration using Skylight orchestrator to automatically measure new service instances while providing continuous 24x7 performance metrics. This is achieved by configuring synthetic and continuous traffic on service paths defined in a 4G or 5G network.

Skylight is designed for distributed and lightweight edge architecture and can be deployed in a fully virtualized environment.

With 5G networks and services, the quality of performance data will not only depend on the granularity, precision, and accuracy of the performance management system, but also on the ability to embed performance management software to dynamically measure network slices and service chains on demand “per customer.”

Analytics are needed to process that streaming data and feed service quality insight in real time into orchestration systems.

Skylight analytics supports closed-loop automation, machine learning, actionable insight through service performance-monitoring visualization, and data-cleaning capabilities using industry-based standards.

Gain visibility to automate assurance of the customer experience

Orchestrated assurance enables you to deliver services right the first time. Maintain customer satisfaction with performance visibility, precise and real-time insights, and the ability to act immediately with Cisco automation.

For more information on Cisco’s network automation portfolio for service providers, please visit https://www.cisco.com/go/crosswork.

For more information on Accedian Skylight, please visit https://accedian.com/platform/skylight/.

To learn more about differentiating your services with exceptional customer experience, or to schedule a demonstration of Accedian Skylight, contact your Cisco sales representative.

 

 

 

Learn more