Skip to content

How jitter, latency, and packet loss
impact your success with unified communications

SupportCallEmailWhether you are a small business or large corporation, chances are you either have already or are contemplating replacing your old POTS/PBX telephone system with some flavor of Voice over IP (VoIP). There are a myriad solutions on the market for almost every organization size. Increasingly VoIP isn’t simply a phone system upgrade but part of a rich unified communications suite, giving your users easy and flexible access to real-time voice, video, screen sharing, and online collaboration.
 

You don’t even have to have a seven-figure IT budget to get these features. Pay-as-you-go SaaS offerings like Skype for Business, GoToMeeting, join.me, BlueJeans, WebEx and others have become essential tools for many organizations. But before you cut the phone cord you need to understand how your network performance (or lack thereof) can make or break your user experience and productivity with these tools.

Quality of Service Metrics for Unified Communications

We’ve all suffered through live meetings or VoIP calls where the audio and/or video was bad. A few extra seconds in a webpage load might go unnoticed but when audio and video is choppy, delayed, or has gaps, it can be excruciating, or at least comical.

To ensure Quality of Service (QoS) for VoIP and unified communications you need monitor and manage three basic network metrics:

  • Latency – how much time it takes for data packets to get from one point to another.
  • Jitter – the variation in in latency over time.
  • Packet Loss – the percentage of packets that never reach their destination.

Let’s look at the effects, causes, and remediation techniques for these metrics.

Latency

Ever been on a conference call with somebody in an adjacent office or cubicle? You end up hearing everything they say twice, first directly from them and (hopefully) through your headset or speakers a short time later.

That’s latency. There’s always some, but when the round trip latency goes above about 250ms we start to notice it – awkward pauses and people talking over each other. Since some latency is out of your control (for calls that traverse the internet) but you should aim for internal network latency levels of less than 150ms.

Latency problems are usually the result from packet handling and queuing delays; essentially packets get bogged down at your network nodes, either because the network is not configured to expedite delivery of VoIP/UC traffic or there is insufficient bandwidth.

Jitter

Most VoIP/UC solutions employ packet buffers to compensate for some amount of jitter, so it’s often not as easy to detect problems on calls. High jitter rates (i.e. large fluctuations in latency) will often result in dropped packets and gaps and audio/video transmission. In general, you want to have internal network jitter of less than 100ms. Jitter beyond that level will result in additional transmission delay (from the buffer processing) and potentially packet loss. Jitter is also often a sign that your network nodes are not configured to prioritize VoIP/UC packets.

Packet Loss

VoIP is highly susceptible to packet loss. Even a 1% packet loss will noticeably affect most real time protocol (RTP) codecs. VoIP/UC clients will recover the best they can but the user experience will be very poor, resulting in lost productivity, aborted calls, and sometimes headsets that are smashed to bits in a fit of rage.

Packet loss is often a symptom of latency/jitter issues, but can also be the result of network misconfigurations, insufficient bandwidth, or simply network equipment that is not able to handle the increased demands of real time communication and collaboration.

Getting Ready for VoIP

To get ahead of these problems, you need to implement a monitoring and remediation plan early, ideally before you deploy these IP communications solution widely. This plan should include the following:

Baseline Testing and Ongoing Monitoring

You can’t fix what you don’t see. It’s a good idea to run some diagnostics/synthetic point-to-point tests within your network to measure baseline latency, jitter, and packet loss metrics so you can see how well your network is configured today. However, it’s also important that you continue to monitor these metrics as you scale your VoIP/UC deployment so you don’t get blindsided as users come online.

Exoprise CloudReady® provides easy use monitoring solutions for standard VoIP/UC as well as Lync/Skype for Business deployments. You can see real time and trend data for your locations and compare that data with aggregate trends measured at other Exoprise customer locations. You’ll be able to see how well your network stacks up to others.

Network Configuration Improvements

Your existing network equipment may support higher VoIP QOS but require configuration changes to do so. Configuring to prioritize VoIP traffic will improve both latency and jitter. Bandwidth reservation, policy-based network management, Type of Service, Class of Service, and Multi-Protocol Label Switching (MPLS) are techniques generally used for prioritizing VoIP traffic.

Network Equipment Upgrades

It may be time to replace older and less capable network equipment. High quality VoIP routers will have a significant impact on these issues and are a must if you are using VoIP as a complete replacement to your legacy phone system. In addition you should consider upgrading your internet connection. You’ll need more speed and bandwidth going forward.

With a solid monitor, reconfigure, and upgrade plan in place your users will have fewer call problems and you’ll have few mangled headsets to replace.

Patrick Carey is VP of Products and Marketing at Exoprise

Back To Top