Businesses that implement voice over Internet Protocol systems in their offices are usually looking to save money. But they certainly don’t want to sacrifice quality. And, on occasion, VoIP calls can be jumpy or suffer from dropped sound. With the implementation of VoIP monitoring, companies can detect problems and look for ways to address them, and using the network approach to VoIP monitoring provides the best results.
In some cases, problems with calls arise from a mistake in the setup, such as a network blockage or a firewall, or in the message itself, when a PBX (News - Alert) does not recognize who you want to call or there are configuration problem. More common, though, are network problems. These occur when network conditions do not permit signaling. These problems might include packet loss or dropped packets, for example, and can have a strong effect on the media side of the call. According to David Roberts, director of product management, enterprise communications, for OPNET Technologies (News - Alert), network engineers who care about VoIP should make sure that the network is adequate to support VoIP applications.
When dealing with a voice call, unreliable transmissions can be a huge problem, since hearing the other person’s words a few seconds later is certainly no good. Whether the issues are caused by high latency -- either via routers that can’t keep up or the simple distance of the call -- or other factors like packet loss or jitter, a poor quality network will impact the call experience.
So, then, what is the best approach to monitoring? Roberts states that the key is to collect information in real time off the network. IT managers can span ports and get feeds from all areas into one stream (for small offices) or several. You can tap into the traffic in each office using span technology and send it to management appliance software and sniff it. This gives you visibility into all media streams for the office, allowing you to see the packets coming in and if there was a delay, for example, or something was missing. You can then apply algorithms to measure the quality of the call and look at network traffic to see what was happening at the time of the call.
With real-time monitoring, you can look into a bad call, learning when it was not good. A report will tell you when it happened, to correlate it with other factors, such as high call volume, or perhaps detect that someone else on the network was using a high-bandwidth application at the time of the bad call. You can drill down to see if some calls were not properly tagged due to misconfiguration, and troubleshoot any issues that may arise.
If the call was of poor quality but you are still not sure why, you can look at other factors. Perhaps there was an issue with the path the voice stream took and the way it was routed. You can look at all the hops and interrogate them. You can eliminate certain factors and see if something else, like a faulty headset, was involved. You can even play back a call as an audio file to listen to the exact problem.
OPNET solutions can be integrated with other network equipment and can report what is happening on the network, centralized to the AppResponse Xpert solution, a single appliance that can help monitor the network and manage the VoIP calls. It is one strong solution to help enterprises.
To find out more about OPNET Technologies, visit the company at ITEXPO Austin 2012. To be held Oct. 2-5 at the Austin Convention Center in Austin, TX, ITEXPO is the world’s largest communications and technology event. OPNET’s Vice President of R&D, Gurmeet Lamba (News - Alert), is speaking during “Migrating Your Apps to the Cloud: Planning and Key Considerations.” For more information on ITEXPO (News - Alert) Austin 2012, click here.
Edited by Brooke Neuman