Home > Articles > Cisco Network Technology > IP Communications/VoIP > VoIP: An In-Depth Analysis

VoIP: An In-Depth Analysis

Chapter Description

This chapter explains many of the issues facing Voice over IP (VoIP) and ways in which Cisco addresses these issues.

Packet Loss

Packet loss in data networks is both common and expected. Many data protocols, in fact, use packet loss so that they know the condition of the network and can reduce the number of packets they are sending.

When putting critical traffic on data networks, it is important to control the amount of packet loss in that network.

Cisco Systems has been putting business-critical, time-sensitive traffic on data networks for many years, starting with Systems Network Architecture (SNA) traffic in the early 1990s. With protocols such as SNA that do not tolerate packet loss well, you need to build a well-engineered network that can prioritize the time-sensitive data ahead of data that can handle delay and packet loss.

When putting voice on data networks, it is important to build a network that can successfully transport voice in a reliable and timely manner. Also, it is helpful when you can use a mechanism to make the voice somewhat resistant to periodic packet loss.

Cisco Systems developed many quality of service (QoS) tools that enable administrators to classify and manage traffic through a data network. If a data network is well engineered, you can keep packet loss to a minimum.

Cisco Systems' VoIP implementation enables the voice router to respond to periodic packet loss. If a voice packet is not received when expected (the expected time is variable), it is assumed to be lost and the last packet received is replayed, as shown in Figure 7-4. Because the packet lost is only 20 ms of speech, the average listener does not notice the difference in voice quality.


Figure 7-4 Packet Loss with G.729

Using Cisco's G.729 implementation for VoIP, let's say that each of the lines in Figure 7-4 represents a packet. Packets 1, 2, and 3 reach the destination, but packet 4 is lost somewhere in transmission. The receiving station waits for a period of time (per its jitter buffer) and then runs a concealment strategy.

This concealment strategy replays the last packet received (in this case, packet 3), so the listener does not hear gaps of silence. Because the lost speech is only 20 ms, the listener most likely does not hear the difference. You can accomplish this concealment strategy only if one packet is lost. If multiple consecutive packets are lost, the concealment strategy is run only once until another packet is received.

Because of the concealment strategy of G.729, as a rule of thumb G.729 is tolerant to about five percent packet loss averaged across an entire call.

7. Voice Activity Detection | Next Section Previous Section