2004 — Yankee Group Outlines the Four Steps Enterprises Should Follow to Manage VoIP

Dec 20, 2005 | Conteúdos Em Ingles

An increasing number of enterprises among all vertical markets have adopted and deployed VoIP across their platforms to achieve different goals, such as cost reduction, improved efficiencies and expanded call features. Yankee Group projects the US retail business VoIP market will reach $3.3 billion by 2010.

Although implementation and use of VoIP may differ, enterprises generally follow the same VoIP deployment path:

· Assessing the need for VoIP

· Evaluating the best provider to fill the previously identified need

· Establishing an implementation plan

First, enterprises need to evaluate whether VoIP will help them achieve their business goals. After identifying and confirming the need, enterprises brace for provider selection and VoIP solution selection. Then the real work starts.

VoIP brings the voice, data and video worlds together in a single pipe. Therefore, enterprises must undertake a dynamic approach to their VoIP solution to bridge the disconnect that occurs between the legacy voice world and the data world. Enterprises must coordinate implementation of VoIP solutions across voice services and the data infrastructure as well as the IT and application environments.

This dynamic approach looks beyond the simple deployment of VoIP. It evaluates the way VoIP affects the rest of the enterprises’ other applications throughout the network infrastructure. As this infrastructure evolves over time, enterprises must follow an approach that takes into account this aspect. The dynamic method is a continual process that helps enterprises to constantly watch and improve their VoIP performance.

The following points illustrate the steps enterprises’ should take:

Step 1: Determine Network Readiness to Deploy VoIP
Most enterprises implement VoIP as a communications environment; they don’t fully leverage its converged nature to carry voice, data and video traffic. Therefore, corporations ought to ensure that voice, data, video and other mission-critical data coexist harmoniously in a VoIP ecosystem. There are three methods to evaluate network readiness for VoIP:

· Method 1: Synthetic call transactions simulate calls across the network and analyze results.

 Pros: Quantifies parameters before deployment

Cons: Lack of existing data traffic affects VoIP availability and quality analysis, and additional overhead

· Method 2: Management information base (MIB) polling polls traffic data from routers and equipment and average utilization every 5, 10 or 15 minutes.

 – Pros: Uses actual traffic data

 – Cons: Skewed results may give a false result of VoIP readiness and subjective measures as some metrics are based on human ratings

· Method 3: Monitoring of detailed user data measures actual (non-simulated) user traffic on an almost real-time basis.

– Pros: Real-time visibility on concurrent calls’ quality and network performance

– Cons: Real-time data collection may affect the network itself, and additional overhead to collect data

Step 2: Identify and Isolate Network Performance Issues
VoIP still suffers from the legacy of a poor enterprise reputation in the 1990s; even though today’s infrastructure provides good quality service. To measure VoIP quality, enterprises use mean opinion score (MOS) as a metric to monitor the performance of each call. MOS provides a numerical indication of the perceived quality of received human speech over the connection. MOS is expressed as a single number in the range of 1 to 5 (1 is lowest perceived quality and 5 is the highest perceived quality). MOS is generated by averaging the results of a set of standard, subjective tests in which a number of listeners rate the audio quality of test sentences read aloud by both male and female speakers over a communications medium. However, MOS is only an indication of poor VoIP performance. Enteprises must then go through a checklist (e.g., inadequate latency, abnormal jitter, lack of bandwidth and compression issues) to determine the reason behind poor performance.

Enterprises should deploy monitoring tools that will give them clear visibility into network performance and the exact root cause of substandard VoIP performance (e.g., introduction of new data applications, issues with providers’ networks, jitter, delay, misconfigured class-of-service settings). Once they identify and categorize service degradation causes, enterprises can isolate the performance problems.

Step 3: Resolve Issues and Monitor Performance
After locating and isolating the reasons for poor performance, enterprises troubleshoot the operating environment before problem resolution. Timely response to problems is vital to enterprises’ business because it directly affects revenue and profit. For example, a large enterprise mentioned that it loses $20 million for every hour of downtime on its network. Therefore, enterprises must troubleshoot and solve these issues before they gain ground and affect additional sites and users.

Real-time monitoring tools are essential to keeping VoIP performance at accpetable levels. Network managers ought to make sure they are equipped with systems to pinpoint real problems, isolate them, troubleshoot them, take appropriate actions to solve them, and then monitor the performance of the VoIP application post-resolution.

Step 4: Look for Trends and be Proactive
Learning from experiences is critical. Building prediction tools that help formulate proactive responses is even more critical. Trending VoIP network usage, requirements and performance constitute a great base of knowledge for enterprises to leverage as they navigate an increasingly uncertain world. So many external factors, such as industry consolidation and site expansion, affect enterprises’ VoIP network on an ongoing basis. As the corporate environment and structure change (e.g., M&A and downsizing), VoIP network readiness may change on a nearly daily basis. Therefore, continual and dynamic management of the VoIP cycle becomes vital for optimized network performance. Incorporating these changes leads to better control over the different variables affecting the VoIP applications and network. Preemptive VoIP control may occur in areas such as traffic prioritization and bandwidth management.

Learning from past experience provides a dynamic and flexible foundation for enterprises’ VoIP implementations. Once deployed, experience will help identify and isolate network performance issues, resolve problems and establish stable and useful performance monitoring, reporting and trending capabilities that regulate variables and enhance performance.

Source: Yankee Group, 2005

Several companies have put together robust network performance management tools for enterprise VoIP deployment that help users better assess and manage their VoIP application and network infrastructure. For instance, Visual Networks (recently acquired by Fluke) created Visual UpTime Select to provide comprehensive visibility from the local loop through individual application flows. Another solution, Micromuse’s Netcool for Voice over IP (NfVoIP) provides a single, business-focused view of IP telephony and data services designed to address the functionality and scalability requirements of networks.

Recommendation for Service Providers
· Show enterprises how to improve network performance and provide them with tools to do so. The success of VoIP business solutions relies heavily on the financial and operational success of a company. There are great virtues in improving companies’ network infrastructure, but you must understand the business challenges companies face and ensure solutions contribute to a smooth process and don’t adversely affect VoIP network performance. However, as you encounter resistance from enterprises when it comes to proving the value of VoIP, you should educate and train them about implementing VoIP. Training and education shouldn’t take the form of a sales pitch—rather, you should present a full view of the benefits, risks and glitches associated with implementing a VoIP solution as well as the impact it may have on network and application performance. A follow-up discussion should cover the ways companies can overcome these glitches and maximize solution benefits.

Recommendation for Software Vendors
· Form sales partnerships with services providers and systems integrators. As service providers become increasingly successful at selling managed services, enterprises will receive such tools from their primary providers, further isolating software vendors such as Micromuse and Fluke. Although software vendors sell through VARs mainly, it might be wise to form partnerships directly with carriers and systems integrators. Under this scenario, software would have two main sales channels: one through VARs and the other through service providers and systems integrators. Software vendors also should put together a training/educational program to help enterprise understand the vitality of such instruments and to help service providers package this service with the larger solution.

Recommendation for Enterprises
· Take a proactive attitude toward preventing VoIP performance issues by implementing your own tools. Don’t rely solely on service providers to monitor the performance of your VoIP network application. Instead, take proactive steps to have your own reporting and monitoring tools. Seeking these tools from companies other than service providers could provide enterprises with an independent and unbiased instrument to measure call performances and prevent network problems.



Notícias – Press-Releases