jump to navigation

VoIP Troubleshooting 101 March 19, 2010

Posted by TelUS Consulting Services in VoIP catagory.


Transmission of voice and transmission of data across an internetwork differ in a number of ways, including the following:

TCP-based data applications react to dropped packets, whereas UDP-based voice applications can only conceal dropped packets. Data applications respond to dropped packets with some degree of correction because they are often TCP-based (TCP resends dropped packets). Voice (which relies on the best-effort transmission of UDP) cannot truly respond to and recover from packet loss, although in some cases the complex algorithms underlying voice transmission can conceal packet loss.

Voice is sensitive to delays, but data is not. Delay-insensitivity means that data applications can tolerate delay well because they are not real-time–based. Voice responds negatively to delay, creating so-called “holes” in the transmission as heard by the receiver.

Basic Requirements for Voice Traffic

Voice traffic is intolerant of packet loss and delay primarily because these conditions degrade the quality of voice transmission delivered to the listener. Delay must be constant for voice applications. The complete end-to-end delay budget for voice traffic cannot exceed 200 milliseconds (ms). As a result, voice traffic passing through a network with congestion or other problems is much more likely to be adversely affected than data traffic passing through the same network.

Common Problems Affecting Voice Traffic

The following factors can affect voice quality:

Delay is the time it takes for packets to travel between two endpoints. Delay manifests itself when long pauses in conversation cause the person listening to start to talk before the other person has finished.

Echo causes a caller to hear the sound of his or her own voice. In most cases, the echo comes back too quickly to be noticed by the caller, but the greater the delay, the greater the likelihood that the echo will be noticeable. (Another important factor is the length of the connection—on international calls, for example, echo is far more likely to be noticeable than it is on local calls.)

Jitter is a variable-length delay that can cause a conversation to break and become unintelligible. Jitter is not usually a problem with Public Switched Telephone Network (PSTN) calls because the bandwidth of these calls is fixed. However, in VoIP networks where existing data traffic might be bursty, jitter can occur.

Latency is the amount of time between when a device requests access to a network and when it is granted permission to send. End-to-end latency describes the overall delay associated with a network.

Serialization occurs when a multiservice route processor (MRP) or router attempts to send both voice and data packets out of the same interface. In general, voice packets are very small (80 to 256 bytes), while data packets can be very large (1,500 to 18,000 bytes). On relatively slow links, such as WAN connections, large data packets can take a long time to transmit onto the network. When these large packets are mixed with smaller voice packets, theexcessive transmission time can lead to both delay and jitter. The time that it takes to put voice traffic onto a transmission line depends on the data volume and the speed of the line—for instance, it takes about 5 ms to send a 1024-byte packet on a 1.544-Mbps T1 line.

Loss occurs when networks drop voice packets. Packet loss is most likely to occur where the network connects to the WAN, although it can occur anywhere in the network.

Noise (or distortion) is a problem that users commonly describe as “muffled,”  “tinny,” or “scratchy.” Noise is typically a result of compression, decompression, packet loss, or echo cancellation.

our next installment : Diagnosing and resolving Echo Problems.

Joe Buck, N.C.E.

%d bloggers like this: