The coming of IP Communications to the world’s networks means that not just service providers and enterprises, for communications equipment manufacturers and communications software applications developers must take into account a complex, hybrid TDM / packetized network and how this impacts on Quality of Service (QoS) and Quality of Experience (QoE). Test applications must deftly handle everything from legacy analog and digital telephony to today’s VoIP and IP video networks. Moreover, the number and variety of devices plugging into this “network of networks” has multiplied as well. To test the quality of real-time IP communications in such an environment, today’s makers of test and measuring equipment and services have devised a multitude of software and hardware including handhelds, portables, benchtops, and rackmount test platforms for field, lab, data center and central office operation.
One interesting company in this area that’s been around since 1989 is Touchstone (News
) Technologies. Their test and measurement equipment and services for communications derive from a talented group of engineers who also happen to lead the sales and management teams. Unlike any competitors, Touchstone provides extensive pre- and post-sales support, taking time to help customers with their evaluation of the products; this can range from web demos to assisting with configurations and testing scenarios.
As for Touchstone’s actual products, the fall into the two major categories of active and passive testing. Active testing products include those for H.323 and SIP Load Generation – in particular, WinSIP software, which turns an ordinary PC into a SIP testing station capable of handling high density loads. WinSIP generates bulk calls, performs feature and function testing, and can test advanced media capabilities. Win WinSIP (and no specialized equipment) a modern mid-tower machine can generate nearly 1,000 simultaneous calls with perfect G.711 audio streams and nearly 2,000 with the G.729 codec. WinSIP is even able to learn new codec types, and offers optional R Factor and MOS scoring for end-to-end quality assessment verification. WinSIP now includes TestML, an XML scripting engine, thus allowing WinSIP to emulate any device and call flow using easy-to-use XML scripts.
In terms of passive testing, Touchstones’ Net Observer can do distributed real-time protocol and media analysis. It’s a scalable carrier-class monitoring and diagnostics solution for voice and video-over-IP communications. Net Observer works by placing software probes near the customer premise and one or more aggregators situated in the provider’s Network Operations Center (NOC (News
)). Net Observer provides more than 125 metrics per call and reports call detail records in real-time. The detail records are stored in an SQL database and the results can be integrated into existing reporting systems through flat files, SNMP or the Radius protocol. Software probes scale in capacity from 4 calls to more than 1,000.
Another Touchstone passive testing solution, WinEyeQ, an extremely accurate voice and video analysis tool, gives you a real-time, top-down display of network activity (in a friendly drill-down format), and delivers up to 650 metrics per call while on-net, 325 per completed call. There are extensive watch and alarming mechanisms and you can capture and record calls and their media for further forensics. You can even import WinPCap trace files.
Touchstone’s TraceBuster utility is yet another passive testing solution, one capable of deciphering SIP and H.323 calls over converged networks. Best of all, it’s free. It gives you a top-down display of call activity, has the Touchstone, drill-down interface, watch and alarming mechanisms, performs detailed signaling analysis, and you can also import WinPCap trace files.
Hammering Out Your Problems
Since the 1990s, Yours Truly has followed Hammer, now called Empirix (News
), one of the best-known names in voice testing and monitoring solutions for network equipment manufacturers, service providers and enterprises. Their expertise extends to contact center applications, unified communications, VoIP, IMS and the NGN. Indeed, back in my days at Computer Telephony magazine, I assembled a testing lab based in part on Empirix’ Hammer equipment. You really get a huge choice of platforms with Empirix. There’s the Hammer NXT for TDM and IP load testing, the Hammer FX for TDM and IP feature testing, and the Hammer UPM RAS and Modem Test for load testing. For VoIP testing, there’s the Hammer Call Analyzer for protocol analysis, the Hammer PacketSphere network emulator for testing IP impairment and the Hammer VQTS to do voice quality testing. For general IP Network testing there’s the Hammer ST (signaling feature and load testing) and the RealStreamer IP for media load tests. The list goes on and on.
Recently, JDSU, itself a leading provider of communications test and measurement solutions and optical products for telecom service providers, cable operators, and network equipment manufacturers, announced that JDSU’s NetComplete service assurance platform now incorporates Empirix’ Hammer XMS, thus creating an integrated, all-encompassing solution for content traffic and signaling-layer monitoring by those service providers wanting to deliver high quality service over broadband/IP wireline and wireless networks. This arrangement between JDSU and Empirix came about partly because service providers are starting to use Ethernet as their primary transport protocol, and such large-frame (packet) networks demand careful analysis of real-time traffic. By adding the Empirix Hammer XMS to the mix, NetComplete can now provide a wide-ranging service monitoring solution that handles both signaling and media content. It also can correlate across multiple protocols and quickly identify network problems, ultimately bringing about a satisfactory customer experience.
Devfoundry Software makes the VoIP Aware VoIP Management system for voice and fax-over-IP deployments. VoIP Aware can be used by VoIP novices as well as experienced administrators. VoIP Aware is a proactive diagnostics tool; a CDR generation utility; and a simple call recorder, all in one package. It identifies and solves problems with their voice and fax over IP deployments by proactively generating diagnostic data for every SIP call in real-time, and continuously monitoring and capturing all SIP calls on a network, thereby eliminating the need for reactive testing and ensuring that administrators have access to all of a call’s original diagnostic data immediately after the problem appears.
Devfoundry’s President Athir Nuaimi, says, “There’s been a renewed interest in fax-over-IP testing. The SIP Forum (News
) has put together a task group to try and get that fast-tracked. T.38 was formulated over a decade ago, but it really never went very far. Many people put a lot of blood, sweat and tears to try to get it to work, and have had limited success in some small domains. But the biggest issue is that the large carriers really haven’t supported it until the last year or two. Global Crossing now supports it, and I believe most of the other major backbone VoIP providers either have T.38 support or it’s in the trial phase. It’s taken a long time, but the stuff is actually now available to customers if they want it, but there’s still some compatibility issues, enough so that a number of what are basically equipment vendors as well as some service providers, have put together this ‘task force’.”
“We were founded about five years ago and originally did consulting work for companies,” says Nuami, “and then about two years ago we came out with our own product. This emerged from some work we did for a fairly small ISP that was working with a large carrier and had issues with VoIP calls. The carrier thought everything was fine, and the carrier, being a well-funded organization, was supposed to have their own quality monitoring systems and have the best test tools. But the stuff just didn’t work. So we created a tool for the ISP that continuously and proactively managed and monitored calls. They basically recorded every single call and anytime there was a problem from one of their corporate customers, they would be able to go back to that call and have the full stats, all the bits and bytes in the call, and be able to go back to both the equipment vendor and the carrier and challenge them on each and every call. The carrier approached the problem merely by setting up some test calls in their testing lab, but it never focused on live calls. Our customer base could have told them that it’s the wrong way to do quality management. You have to do it with live traffic, and you have to do it on an entire set of calls. You can’t sample a certain percentage of them. So we had a very aggressive approach, but that proactive way of dealing with it actually meant that they could tell exactly where the problems were very quickly, and there was no retesting required. I personally believe that that experience showed me that’s the only way to manage quality on a VoIP network. You can’t do it reactively; you can’t do it in a ‘testing place’. You must do it on live traffic, and all the time.”
Nuami elaborates, “If you were to call your phone company – next-gen or otherwise – and told them that yesterday’s calls were pretty bad, I’d be shocked if most of the phone companies would even be able to look at those calls and say, ‘You’re right, yesterday 37 percent of your calls really didn’t meet our standard for the quality that we want to offer you, and as a result, we’re going to give you a rebate on yesterday’s calls. And we believe that, whatever the problem was, we’ve fixed the network and it won’t happen today’. I don’t know of a single carrier that offers that kind of service. But that’s really what a proactive approach is all about; that the carrier should be able to look at your traffic from yesterday and say, ‘You’re correct, it’s bad. You just called us, but I already have the stats in front of me.’ And that’s our approach to testing. We look at all the traffic and we do the analysis on all the calls, and we do it all day, every day. All the data goes into a database, and the stats are there for the customer as they need it. The information could be as simple as ‘Was the call of good quality or not?’ or it could be the entire capture for the call that can be sent to Engineering or equipment vendors to diagnose.”
Richard Grigonis is Executive Editor of TMC�s IP Communications Group. To read more of Richard’s articles, please visit his columnist page.
Edited by Stefania Viscusi