As an IT administrator, engineer, or support specialist, you know how important the performance of Microsoft Teams is to your business. When Microsoft Teams is glitching or crashing, collaboration (which is why you use Teams in the first place) is difficult or impossible for your users. When that happens, they turn to you. Troubleshooting network issues is often the first step to solving the problem (or at least proving that the network is not the problem). This blog post provides an overview of Roundtrip Time (RTT), Jitter, Packet Loss, and Bitrate and the ideal values for each during a Teams call or meeting.
It’s difficult to define a perfect value for network metrics that is true in every case. New features are constantly released, and user expectations are always increasing. Given the pace of change, a “fast” connection as we write this might only give you “meh” performance when you read it. Posts like this can quickly seem laughably out of date and come back to haunt the writer.
With that disclaimer, there are some generally accepted values for what constitutes a “good” connection for Microsoft Teams calls and meetings today.
Roundtrip Time (RTT): 300 ms or lower (150 ms is ideal)
RTT measures the time it takes for a data packet to travel from point A to point B and back. It’s usually measured in milliseconds. Consistently higher RTT values indicate network latency issues. When you see people unintentionally talking over one another in calls, it suggests that RTTs are too high.
Jitter: 30 ms or lower
Jitter is another useful metric. While RTT measures average latency in milliseconds, Jitter measures the variability of that latency. Low RTT plus high Jitter often leads to intermittent issues. You can have a great connection on average, but if latency goes through the roof for a few seconds every once in a while, you can have problems. Spikes like these can cause intermittent issues with call quality, like choppy audio streams.
Any Jitter value over 30 ms should be a concern. The problem could be caused by undersized network hardware, but it could also indicate that your Teams-related data is competing for network resources with other data and losing.
Packet Loss: below 5%
As you know, on any ethernet network using TCP, the sending device breaks your data down into a series of smaller packets, numbers them, and sends them to the receiver. Because it was developed for the public internet, where each packet might follow a different path to the destination, data packets can arrive at the receiving device out of order. The receiver puts the packets in the correct order before further processing occurs. But when one of the packets goes missing, the receiving device tells the sender to start over again at the beginning.
High packet loss can result in missing audio and video data, resulting in poor call quality and frozen video. This issue is often a sign of a poor connection, failing hardware, or insufficient bandwidth. Video streams are particularly sensitive to packet loss.
Bitrate: 50-100 kbps for audio, 1.5 Mbps or higher for HD video
Bitrate requirements can vary widely depending on the workload. Video resolution is the perfect example of the idea that today's state-of-the-art becomes tomorrow's expectation. A bitrate of 1.5 Mbps or faster is acceptable for HD video, but 4 Mbps or faster is recommended. Monitoring the bitrate ensures that enough bandwidth is allocated for audio and video transmission. Inconsistent or insufficient bitrates can result in reduced video quality and audio dropouts.
Quality of Service (QoS) technology categorizes and marks different types of network traffic and configures routers to process them based on priority. This lets you reserve bandwidth for things like voice and video regardless of how much other traffic is traveling the network, as shown in Figure 1. Ensuring that Microsoft Teams traffic always has sufficient bandwidth will go a long way toward eliminating issues.
Figure 1. Network traffic prioritization by category showing reserved capacity for voice and video data.
Many networks are intentionally over-subscribed, meaning that total network capacity is less than the maximum theoretical demand of everything connected to the network. It assumes that only a few devices will max out at a given time. This is usually a safe assumption, but as the number of devices on the network grows, it could become a problem. If your network is congested because more devices are connected than it was designed to handle, increasing the capacity of your network hardware and connectivity can solve the problem.
Understanding what “good” looks like when interpreting networking metrics for Microsoft Teams can help you diagnose network issues and provide solutions more quickly. (You can also show that the network is not the source of the problem, so you can start looking elsewhere.)
Fusion Connect can help you get the most out of Microsoft Teams, including tuning your network for voice and video. We can also help you add more advanced features to Teams, like texting (SMS/MMS), automated attendants, contact center solutions (CCaaS), call recording, and AI-powered speech-to-text transcription. If you haven’t deployed Microsoft Teams yet, we can help you assess your existing network and identify areas that might be a concern.
If you would like more information on this topic, you might find these resources interesting: