Service Performance Assurance Solution for LTE

Cisco 1 estimates global traffic from mobile broadband will grow at 92 percent CAGR between 2010 and 2015. Mobile data services are also gradually overtaking voice as the predominant source of revenue.

Finally, market research firm Ovum predicts that by 2015, the ARPU of mobile data services will pass that of voice services in North America. In response, MNOs are positioning themselves for the future by building out additional network capacity through the transition of their 2G and 3G networks to 4G LTE. The Global Mobile Suppliers Association (GSA 2 ) reports there are already more than 237 operators in 85 countries investing in LTE. The success of these transitions to LTE will be determined by MNOs’ ability to deliver a high-level Quality of Experience (QoE) to the subscriber. The transition to LTE is much more than an upgrade in mobile access technology; it is a transformation of their entire network infrastructure. By unifying core and access networks, LTE flattens the mobile network architecture, improvg scalability and lowering overall operating costs. However, it also forces mobile operators to move away from the familiarity of dedicated-circuit Backhaul networks to a packet-friendly infrastructure like Carrier Ethernet. This move to an all-IP network infrastructure presents new network connectivity and mobile service quality monitoring challenges that MNOs must address in order to minimize the negative impact on the quality of their mobile services. Mastering these challenges will require a new approach to network monitoring and assurance practices. Because mobile operators’ future revenue and growth will be driven by content and application-based mobile services, subscribers’ QoE will determine an MNO’s ability to retain customers and upgrade them to these new service offerings. The widespread adoption of LTE means these services will eventually be deployed over an Evolved Packet Core (EPC) network infrastructure. With this new reality, MNOs must take a fresh look at the tools and processes used to monitor and manage their end-to-end mobile service quality.

MNOs face significant challenges in ensuring subscriber QoE for LTE-enabled service classes and applications. These challenges stem from the difficulties inherent in deploying synchronous applications over packet-based networks and are exacerbated by lack of experience with managing shared-bandwidth data services on an evolved IP access and mobile core network. 

Dedicated-Circuit Reliability versus IP Flexibility

Operators are familiar with dedicated-circuit SDH/SONET-powered mobile infrastructures and, while expensive to deploy, grow and maintain, these legacy technologies are extremely reliable and predictable. In 2G and 3G deployments, mobile operators were able to rely on TDM backend networks to deliver mobile traffic, and therefore to focus their attention almost entirely on the RAN and service quality. LTE’s flat IP architecture leads to a lower deployment cost-per-megabyte, better spectrum management and efficient scalability. IP is an asynchronous packet-based protocol supporting many types of traffic and different conversations that flow across the network simultaneously. By default, IP packets travel in unpredictable paths across heterogeneous networks of physical and virtual elements with no guarantee of when or if they will arrive.

Because LTE ensures low latency and multiple classes of services, subscribers are able to stream high-definition video and TV, play multimedia-rich interactive games and access near real-time business and consumer applications as these synchronous applications do not tolerate the higher latency, jitter, and packet loss of an IP network. Network performance problems, such as poor connectivity or traffic throughput, can significantly affect service quality. Therefore, assuring the performance of these applications will require constant monitoring of network elements and traffic not only in the E-UTRAN and EPC, but also in the underlying IP transport network. 

The initial deployment of LTE will be in conjunction with existing 2G and 3G networks and will extend the scope and complexity of monitoring mobile networks and services. With the introduction of new “4G” RAN and core entities such as eNodeBs, S1 interfaces, Mobility Management Entities (MME), Signaling Gateways (SGW) and Packet Network Gateways (PGW), MNOs will need to assure Ethernet Virtual Connections (EVC), VLANs and Pseudowires – critical entities of the new Carrier Ethernet backhaul infrastructure carrying the traffic to and from cell towers. Ensuring subscriber QoE requires identifying and inventorying all network entities responsible for delivering services. The first challenge for MNOs is capturing consolidated service-level data as utilizing vendor-specific management applications to consolidate cross-domain service-level performance metrics can be inefficient. Similarly, mobile network probes can help, but may miss key performance perspectives such as cell coverage quality, E-Utran scheduler latencies or number of active UEs per node. Mining, correlating and transforming raw data into actionable service quality metrics is also challenging. Turning vendor specific LTE raw counters into actionable network health and performance dashboards requires a multi-vendor tool to create KPI by easily referencing vendor counters.

The initial deployment of LTE will be in conjunction with existing 2G and 3G networks and will extend the scope and complexity of monitoring mobile networks and services. With the introduction of new “4G” RAN and core entities such as eNodeBs, S1 interfaces, Mobility Management Entities (MME), Signaling Gateways (SGW) and Packet Network Gateways (PGW), MNOs will need to assure Ethernet Virtual Connections (EVC), VLANs and Pseudowires – critical entities of the new Carrier Ethernet backhaul infrastructure carrying the traffic to and from cell towers.

 

Ensuring subscriber QoE requires identifying and inventorying all network entities responsible for delivering services. The first challenge for MNOs is capturing consolidated service-level data as utilizing vendor-specific management applications to consolidate cross-domain service-level performance metrics can be inefficient. Similarly, mobile network probes can help, but may miss key performance perspectives such as cell coverage quality, E-Utran scheduler latencies or number of active UEs per node. Mining, correlating and transforming raw data into actionable service quality metrics is also challenging. Turning vendor specific LTE raw counters into actionable network health and performance dashboards requires a multi-vendor tool to create KPI by easily referencing vendor counters.

 

Figure 1 – Cross-domain LTE Performance Counters Collection

 

Visualizing Services End to End 

The multi-technology nature of mobile networks has promoted a silo-oriented service assurance practice amongst most mobile networks operators. Each functional group uses their own set of network data and has developed tools and processes to address the associated service management needs. 

 

With LTE’s flat network architecture, mobile service quality spans across a new set of IP-driven networking domains. Events that affect network performance will likely be caused by a combination of anomalies across the network. For instance, an E-UTRAN engineer who sees a high number of RRC connection drops at a cell site might suspect a network resource availability issue or a saturated sector. But in an LTE network, this identified issue in the access network could be related to a saturated S1 interface, an unavailable VLAN or perhaps a saturated EVC connection to the EPC. Troubleshooting these types of issues will become increasing complex if operators are not able to take a step back and see end-to-end service quality across the entire IP service-delivery infrastructure.

 

InfoVista’s Service Performance Assurance Solution for LTE

A successful LTE transition requires a field-proven Service Quality Assurance Solution that applies IP Core network management expertise to LTE’s EPC and E-UTRAN. At the base of InfoVista’s carrier-grade solution is a fault-tolerant data collection layer that connects to a flexible consolidation engine with intelligent service performance analytics. It builds on these core data processing capabilities with a rich presentation layer comprised of Web 2.0 live reporting interfaces and dashboards. To assure its relevance to end-users, it provides end-to-end, role-based visualization of mobile network service quality, enabling MNOs to effectively manage multiple large-scale network domains from a single platform. InfoVista’s solution assures delivery across multi-vendor IP driven mobile networks and can accelerate transition to LTE by ensuring technology cost savings, operational effi ciency and high subscriber QoE

 

Carrier-Class IP Platform

 

Intelligent KPIs/KQIs 

The value of a service performance assurance solution lies in its out-of-the-box set of Key Performance Indicators (KPIs), Key Quality Indicators (KQIs) and the intelligent algorithms and presentation layers that turn raw network data into actionable service-oriented intelligence. Years of experience in assuring IP services performance for mobile networks have helped us build a solution with an extensive set of pre-defined indicators that can be easily tailored or modified by RAN, transport and service quality engineers for their efforts of assuring performance and QoE. 

 

Figure 2 – Examples of Out of the Box KPIs

 

Adaptive Performance Metric Thresholding 

Each mobile network is different. Advanced IP performance measurement technologies, such as intelligent Adaptive Performance Metric Thresholding, establish baseline performance levels for normal network operations throughout a typical day. By identifying what is “normal,” the solution can proactively alert RAN staff of abnormal performance and quality of service issues. This feature helps operators detect and stop network degradation and attacks before the network suffers an extensive impact across the service delivery chain.

 

 

Time Series Analytics 

 

Mobile data service traffi c is “bursty” by nature. It is critical for network and performance engineers to understand how service quality is affected during the most stressful network conditions. Advanced features like Time Series Analytics include monthly reports with Busy Hour/Busy Day, 95th percentile, and standard deviation analyses on performance indicators such as UL/DL throughput and aggregated latency. Through built-in algorithms, reports allow engineers to define the performance of their mobile network when it is most stressed, according to user-defined criteria (busy hour determiners) or user defined time-ranges (i.e. weekly, monthly, quarterly, and yearly). This information helps network engineers make sense of time-variant mobile performance data and supports more informed decisions in day-to-day infrastructure planning and optimization activities.

 

Figure 3 – KPIs and Thresholds Configuration Dashboard

 

Automated Discovery 

 

InfoVista’s Service Performance Assurance Solution for LTE is built on a platform with out of the box support for leading mobile and IT equipment vendor equipment and management systems. Automated engines discover the entities and relationships amongst physical and virtual assets, gather configuration data and display network connectivity and topology structures graphically for all domains. Once the initial discovery is completed, the solution immediately begins monitoring the network infrastructure to collect performance metrics and compute service-level quality indicators. An installation can be functional within hours and no network inventory database or topology spreadsheet integration is required. 

 

Open Collection and Reporting for Mobile Networks 

 

Mobile infrastructure equipment ’s Element-level Management Systems (EMS) can be goldmines of E-UTRAN and EPC performance data. However, each vendor supplies its own set of thousands of raw counters and proprietary metrics. Turning these data into useful, vendor-agnostic performance indicators can be cumbersome, time-consuming and costly. For more than a decade, InfoVista has been developing a robust and scalable Open Collection and Reporting System for today’s mobile network performance management requirements that allows mobile operators to directly reference vendor counters to create actionable multi-vendor KPIs within a matter of minutes, and use them to measure mobile service quality across multiple EMS and domains (RAN to the core). 

 

Selective Configuration of EMS Raw Counter Storage 

 

Many EMS’ raw counters are never used or are unnecessary for the purpose of measuring service quality and performance. Plus, storing this raw counter data adds to IT costs and limits the frequency and granularity at which data is polled. InfoVista’s solution is capable of handling the storage of raw counter data, but it also supports selective processing and the storage of specific raw counter groups. As another extreme, an operator could choose to not store ANY raw counters; but only synthesized KPIs/KQIs. New vendor counters are also picked up automatically without requiring a software upgrade. Conversely, EMSs that do not store volumes of unwanted data may be able to reduce the solution’s total cost of ownership by aligning it with business CAPEX and OPEX reduction strategies. On-Demand Performance Reports

 

 

Figure 4 – Service and Infrastructure Visualization Dashboard

 

Figure 4 – KPIs Analytics Dashboard

 

On-Demand Performance Reports

In addition, users can define “Query Time” KPIs through an intuitive GUI. “Query Time” KPIs are computed on-demand at the time of reporting and are needed to build higher order KQIs, testing formulas and understanding mobile network service quality without impacting the platform sizing. This feature allows RAN, transport and core engineers to dramatically improve operational effi ciency and mean time to repair (MTTR). LTE Network Service Quality Visualization

Figure 5 – Over-time KPIs Performance Dashboard

Service Quality Visualization builds on expertise in each network domain and provides a unified, end-to-end view tailored per functional group and adapted to relevant adjacent domains (i.e. transport and access) for enhanced service-level insight. This collaborative, single pane of glass approach eliminates the traditional performance management silos, resulting in an operational and management practice that improves troubleshooting, reduces MTTR and enables proactive network planning and QoS processes. 

 

End-to-End Network Availability Monitoring Use Case 

Continually monitoring and verifying connectivity is critical to ensure resource availability and minimize service disruptions. Real-time views and performance data at and across network domains help network operators guarantee end-to-end network connectivity and enable comprehensive troubleshooting. For example, when a threshold violation trigger alerts a RAN engineer of a poor timing/synchronization state of a cell tower, the technician can immediately detect and associate this timing/synchronization issue with unusually high latency in the backhaul circuit with a graphic dashboard displaying the E-UTRAN and the associated mobile backhaul in his service-area. Further, the technician can assign the issue to a particular segment in the backhaul by zooming into CRC error due to a faulty card in the transmission network.

 

Figure 2 – EPS Bearer Service Architecture

 

Summary

 

InfoVista is the industry leader in IP Service Performance Assurance solutions for service providers. The company’s carrier-grade solution for LTE helps mobile operators to manage the challenges of deploying and optimizing E-UTRAN and EPC entities and assure service quality over an all-IP network infrastructure. 

 

With experience in all-IP service performance assurance, intelligent and actionable out-of-the-box KPI and KQIs and advanced network performance analysis features like Adaptive Performance Metric Thresholding and Time Series Analytics, InfoVista’s solution provides mobile network operators with unparalleled visibility into the performance and health of their 4G infrastructure. Its scalable and open collection and reporting capabilities support user-defined vendor-agnostic KPIs, flexible “Query Time” KPI reports and selective EMS raw counter storage features, lower the solution’s Total Cost of Ownership (TCO), and improve mobile operators’ operational efficiency. 

 

The “single pane of glass” visualization supports real-time collaboration between functional groups breaking down management silos and speeding resolution of cross-domain service-affecting problems. By providing a single solution capable of monitoring and assuring an entire network infrastructure, from the RAN to the Core, InfoVista empowers mobile network operators to optimize IT costs, better assure subscriber quality of experience and ensure a successful transition to LTE.

 

 

 

AUTHOR SUMMARY

Infovista
www.infovista.com  

InfoVista is the leading provider of service performance assurance software solutions for IP-based network and application services. We empower communication service providers and large IT enterprise organizations to transform their IT infrastructure into a distinctive asset for revenue generation, customer loyalty and business agility by adopting a quality centric approach to expedite the launch of innovative, differentiated and performing services ahead of the competition.

InfoVista’s unified network performance management and application performance management platform equips 80% of the world’s largest operators and a roaster of global enterprises with the actionable visibility they need to ensure a high-quality user experience end-to-end, by holistically and effectively assuring the performance and quality of their converged network and IT services, while keeping operational costs as low as possible. InfoVista is traded on the Euronext Paris (FR0004031649) and can be found online at http://www.infovista.com

 

 

 

For more information, please visit www.infovista.com 

For sales inquiries please email: marketing@infovista.com