November 2007 | Volume 10/ Number 11
Testing IP Communications, Inside and Out
By Richard “Zippy” Grigonis
Interaction among devices in a network topology determines service quality. Network elements are wildly different in nature: IP phones, soft phones, routers, switches, firewalls, gateways, session border controller, applications session controllers, etc. Therefore, the testing of individual devices is only useful up to a point. Thousands of hours of system testing is necessary, involving potentially hundreds of device configurations, device features and network configurations that comprise today’s quadruple play and Next-Gen Networks (NGNs). Pre-deployment testing and automation helps.
For 60 years, Tektronix (http://www.tektronix.com) has been one of the great names in test, measurement, and monitoring products, solutions and services for the communications, computer, and semiconductor industries.
Yours Truly managed to catch Tektronix as it was about to release some news concerning its VoIP and IMS (IP Multimedia Subsystem) hardware and software diagnostic solution for VoIP and converged networks, Spectra2. Spectra2 can monitor, test, generate (simulate up to 60,000 user agents), and automate testing of VoIP and PSTN (SS7 ISUP) signaling and media.
Mike Erickson, Function and Load Diagnostic Product Marketing Manager, says, “Spectra2 is a single PC-based platform running Windows with lots of applications built into it. It’s designed to fit into all phases of a carrier product lifecycle with an emphasis on load generation, functional testing, monitoring and media Quality of Service [QoS]. One area of focus here is protocol depth. What’s special about Spectra2 is that we haven’t abandoned our great PSTN testing capability: ISDN, TCAP, SS8, ISUP - we still have a very strong focus on those, because 95 percent or more of phone calls still traverse the PSTN at some point. So we continue to offer a truly comprehensive test product. And of course we’re very deep into VoIP, with our protocol test capabilities including SIP, H.323, Megaco/H.248, MGCP, and voice codecs and so forth. We’ve taking into account mature technologies, evolving technologies and now, going forward, the IMS infrastructure.”
“In October we announced a new high load platform - it gives us a 700 to 800 percent increase in capacity over our current platform in terms of SIP processing capabilities,” beams Erikson.
Perhaps the most difficult forms of testing involve Fixed-Mobile Convergence (FMC) handover between cellular and WiFi.
At Azimuth Systems (http://www.azimuth.net), a major provider of wireless data communications test solutions, Graham Celine, Senior Director of Marketing, says, “We deal with VoIP purely from a WiFi perspective. But we’re now seeing a much more serious player in the service provider who is putting his business on the line, because he must deliver an offering that has high QoS and is reliable. With FMC, everyone expects a phone call to drop off a cellular phone and handover onto the local WiFi network, but a problem can arise if it’s briefly jumping onto a low-quality network. The trend we’ve seen is that some extended level of testing is done by service providers, but they’ll say, ‘Well, we can provide you with VoIP-over-WiFi as long as you’re doing this and that…’ There is a tendency for some service providers to actually offer the access point for the WiFi cards, and they say that they can’t guarantee that the equipment will work with any other technology.”
Celine elaborates: “The three key factors to test in such environments are: First, performance. I mean more QoS than throughput, since voice doesn’t take up a lot of bandwidth. Second, is range. A WiFi phone system may work if you’re standing next to the access point and doesn’t when you stand 100 feet away, and vice versa. There should be a general good design of the product with voice quality and reliability over range.”
“A third aspect is capability; particularly with FMC-types of VoIP services. A significant amount of handoff goes on in such systems: cellular-toWiFi or WiFi-to-cellular. What’s going to happen when I get home, get out of my car, and walk into my house, or get out of my car in the morning and go to the office? The handover must work.”
The Fanfare Group (http://www.fanfaresoftware.com) is a vendor of innovative testing solutions enabling network equipment makers to simplify and accelerate their device testing. For example, the FanfareSVT test automation software is used by companies such as Juniper Networks for optimal product testing and quality assurance. Also, Fanfare’s iTest package provides quick, easy device testing that’s optimized for feature testers and developers. This is important because device manufacturers must increase their quality to “carrier class” to meet service provider expectations, so extensive device and system testing are required. Moreover, service providers must not only validate device operation but perform comprehensive system testing, which can be complicated and expensive.
Glenn Jones, Vice President of Engineering for Fanfare, says, “Our third-generation application provides a testing and a test automation framework for high tech equipment makers. It’s focused on simplifying and accelerating testing at both the device level - a router, hub, switch or a telecom switch - and also at a system level where you have many devices chained together, including traffic generators and things of that nature. Our market has mainly been in quality assurance [QA], but some developers also use our products for unit testing and some system level testing. What’s key is that we try to enable the testing and QA teams to get their product to market fast. By providing this IDE-type environment - an integrated environment for testing - developers in QA teams are much more effective at building things. They construct the tests and those are executed, recorded and reported, and then the tests can be quickly automated.”
“Testing” Becomes Monitoring and Service Assurance
Yours Truly has always maintained that in a modern, dynamic network, “testing” as such is only viable if it’s reengineered as round-the-clock monitoring.
This is particularly true when it comes to IP Video. Take for example, IneoQuest’s (http://www.ineoquest.com) IQPinPoint, an advanced quality and service assurance solution for IP Video. Service providers employ it to improve video quality and control IPTV operational expenses (OPEX). IQPinPoint can predict, detect, isolate and resolve faults in live IPTV networks via its combination of real-time remote troubleshooting with management of hundreds to thousands of live simultaneous video programs.
IneoQuest’s Tom Tucker, Director of Marketing, says, “Years ago, when digital technology really got going, testing seemed to always be an afterthought. Digital broadcast and related customers thought that there was a sort of ‘cliff’ phenomenon: something worked perfectly to a certain point; once it reached the ‘cliff’ it stopped working. With the development of and access to IP networks, to a large degree that same mentality played through. Network component vendors gave that same impression: you plug the components together, test it in a lab, then you put the content in one end and it either comes out the other end in perfect condition or it doesn’t come out at all. Of course, that’s not really the case, particularly in a packet-switched network. The technology is so complex and the dynamics of an IP network can be such that it really requires 24x7 monitoring on all the streams. Without a doubt, customers are now embracing the idea that they must design-in a service assurance strategy from the beginning of their projects. They’ve learned this the hard way, deploying components and expanding a network and its streams until finally problems appear and the troubleshooting time becomes a killer.”
And Now, Pre-Deployment Testing
OPNET Technologies (http://www.opnet.com) supplies advanced management software for networks and applications: IT Guru Network Planner and SP Guru Network Planner, which automate the analysis and planning of multi-layer, multi-vendor networks. Organizations can use these to more accurately plan for growth and technology migration, optimize network capacity and performance, and accelerate new application and service deployment. The optional NetDoctor module enables users to validate network configuration changes before deployment. A new IPv6 Planning and Operations module is used to plan the migration of an organization’s IPv4 networks to IPv6.
OPNET’s Senior Vice President of Model Research and Development, Pradeep Singh, says, “OPNET plays a pretty unique role in the VoIP market. Typically everybody wants to save money by going to VoIP, but people don’t realize that if they don’t do a pre-deployment analysis, chances are the network isn’t even ready for VoIP. A real risk of failure exists unless certain tweaks are made to the network, such as specific device configurations, or adding redundancy, or adding capacity at various strategic locations. It all makes a difference when a pre-deployment assessment for VoIP is performed.”
“Our approach to the VoIP space is to be the strongest player in terms of providing a readiness assessment for VoIP deployments,” says Singh, “which includes device configuration validation and performance validation. Our customers want to do two kinds of testing when deploying VoIP: First, some active testing, where they deploy some agents and sample VoIP sites, and then they generate some traffic and test for quality across the network in terms of the usual things: latency, jitter, packet loss, the computing of MOS [Mean Opinion Scale] scores or R Factors, and trying to figure out everything is acceptable or not. Secondly, there are ‘passive’ tests where you create a virtual test bed of your network over which you wish to run VoIP and you then perform pre-deployment analyses of that network. That approach aligns very well with our product suite because our first step in any analysis is to create a virtual representation of someone’s operational or production network. By virtual representation I mean that, within software, we create a very detailed, high fidelity network model of the switches, routers, gateways and end point devices communicating over the network, importing the baseline traffic, and representing what kind of services and applications are running in the network. This baseline representation is done before the actual deployment, so you can see what would happen if, for example, you add VoIP across two particular sites.” IT
Richard Grigonis is Executive Editor of TMC’s IP Communications Group.
Today @ TMC
ITEXPO West 2012
October 2- 5, 2012
The Austin Convention Center
The World's Premier Managed Services and Cloud Computing Event
Click for Dates and Locations
Mobility Tech Conference & Expo
October 3- 5, 2012
The Austin Convention Center
Cloud Communications Summit
October 3- 5, 2012
The Austin Convention Center
Jive Communications by LogMeIn has Signed on as a Platinum Sponsor for 2019 Collocated Events, ITEXPO and MSP Expo