Cisco VoIP - Ronak

8/8/2019 Cisco VoIP - Ronak

1/26

Cisco SystemsVoice over IP Technology

Course: VoIPDone By: Ronak S Aswaney, ID:0710229.

Date: 02/03/10.

1


2/26

Overview

VoIP: An In-Depth Analysis

In this Presentation we will highlight the mainissues facing Voice over IP (VoIP) and ways inwhich Cisco addresses these issues.

2


3/26

VoIP: An In-Depth Analysis

We will explain each of the issues in detail and how they can

affect packet networks. Below are the following issues whichwe will cover:

Delay/Latency

Jitter

Pulse Code Modulation (PCM)

Voice Compression

Echo

Packet Loss

Digital-to-Analog Conversion

Tandem encoding

3


4/26

Delay/Latency

VoIP delayorlatencyis characterized as the amount of time it

takes for speech to exit the speakers mouth and reach thelisteners ear.

Three types of delay are inherent in todays telephonynetworks:

Propagation Delay: caused by the length a signal must travel via light

in fiber or electrical impulse in copper-based networks.

Handling Delay: defines many different causes of delay (actual

packetization, compression, and packet switching) and is caused bydevices that forward the frame through the network.

Serialization Delay (Processing Delay): is the amount of time it takes toactually place a bit or byte onto an interface.

4


5/26

Propagation Delay:

Light travels through a vacuum at a speed of 186,000 miles per

second.

Electrons travel through copper or fiber at approximately

125,000 miles per second.

A fiber network stretching halfway around the world induces a

one-way delay of about 70 milliseconds. Although this delay is almost unnoticeable to the human ear,

propagation delays in combination with handling delays cancause noticeable speech degradation.

5


6/26

Handling Delay:

Devices that forward the frame through the network cause

handling delay.

Below I have described handling delays in Cisco Devices andhow they affect voice quality:

In the Cisco IOS VoIP product, the Digital Signal Processor (DSP)

generates a speech sample every 10 ms when using G.729. Two ofthese speech samples are then placed within one packet. The packet

delay is, therefore 20 ms. An initial look-ahead of 5 ms occurs whenusing G.729, giving an initial delay of 25 ms for the first speech

frame. Cisco IOS enables users to choose how many samples to put into

each frame.

*G.729 is an audio data compression algorithm for voice that

compresses digital voice in packets.

6


7/26

Queuing Delay:

A packet-based network experiences delay for many reasons,

out of which two major reasons are described below:

Necessary time to move the actual packet to the output queue(packet switching).

Queuing Delay.

When packets are held in a queue because of congestion on anoutbound interface, the result is queuing delay. Queuing delay

occurs when more packets are sent out than the interface canhandle at a given interval.

The actual queuing delay of the output queue is another cause

of delay.

The ITU-T Standardization Sector recommendation specifiesthat for good voice quality, no more than 150 ms of one-way,

end-to-end delay should occur. 7


8/26

Queuing Delay:

8

Cisco VoIP implementation, two routers with minimal network delay

(back to back) use only about 60 ms of end-to-end delay. This leaves upto 90 ms of network delay to move the IP packet from source to

destination.

Some forms of delay are longer, although accepted, because no otheralternatives exist. In satellite transmission, for example, it takes

approximately 250 ms for a transmission to reach the satellite, andanother 250 ms for it to come back down to Earth. This results in a totaldelay of 500 ms.


9/26

Jitter

Jitter is the variation of packet interarrival time. Jitter is one

issue that exists only in packet-based networks.

While in a packet voice environment, the sender is expected toreliably transmit voice packets at a regular interval (for

example, send one frame every 20 ms). These voice packetscan be delayed throughout the packet network and not arrive

at that same regular interval at the receiving station (for

example, they might not be received every 20 ms; Thedifference between when the packet is expected and whenit is actually received is jitter.

9


10/26

Jitter

You can see that the amount of time it takes for packets A and

B to send and receive is equal (D1=D2). Packet C encountersdelay in the network, however, and is received after it isexpected.

Note that jitter and total delay are not the same thing,

although having plenty of jitter in a packet network can

increase the amount of total delay in the network. 10


11/26

Pulse Code Modulation

Analog communication is ideal for human communication, butanalog transmission is neither robust nor efficient atrecovering from line noise. In the early telephony network, when analog transmission was

passed through amplifiers to boost the signal, not only was the voiceboosted but the line noise was amplified, as well. This line noiseresulted in an often-unusable connection.

It is much easier for digital samples, which are comprised of 1 and 0bits, to be separated from line noise. Therefore, when analog signalsare regenerated as digital samples, a clean sound is maintained.

That is why, we have to convert the analog sound into digitalform;Which is done by the PCM.

PCM - The process of converting an Analog signal to a Digital signalor vice versa.

11


12/26

Voice Compression

Transferring voice data over the internet isn't appropriate,

especially if kept in the same state. It will be too heavy andcomplex to transfer the voice data. This is why we need to

compress the voice data into a different format which willbe lighter and easier for transfer.

Voice Compression is a process whereby voice data iscompressed to make it less bulky for transfer.

Compression software (codec) encodes the voice signals into

digital data that it compresses into lighter packets that are thentransported over the Internet.

At the destination, these packets are decompressed and giventheir original size, and converted back to analog voice again, sothat the user can hear.

12


13/26

Voice Compression

Two basic techniques are used for Voice Compression:

PCM

a-law

-law ADPCM - Adaptive differential pulse code modulation.

In PCM, a-law & -law are similar, they both use logarithmiccompression to achieve 12 to 13 bits of linear PCM quality in 8

bits. Although -law has a slight advantage over a-law regarding in

the SNR performance.

ADPCM encodes using 4-bit samples, giving a transmissionrate of 32 Kbps.

PCM and ADPCM are examples ofwaveform codecs.

13


14/26

Voice Compression

Codec Bandwidth/kbps Comments

G.711 64 Describes the PCM voice coding technique outlinedearlier. Voice is already in the correct format for digitalvoice delivery in the public phone network

G.726 16/24/32/40 Describes ADPCM coding.You can interchangeADPCM voice between packet voice and public phone,provided that the latter has ADPCM capability.

G.729 8 Excellent bandwidth utilization. Error tolerant. Licenserequired.

G.723.1 5.3/6.3 High compression with high quality audio. Can use withdial-up. Lot of processor power.

iLBC 15 Robust to packet loss. Free

14

The most popular voice coding standards for telephony and packet

voice include:


15/26

Echo

It is the repetition of a sound resulting from reflection of the

sound waves.

Echo on a phone conversation can range from slightly

annoying to unbearable, making a conversation unintelligible.

Echo has two drawbacks:

It can be loud, and it can be long.

The louder and longer the echo, of course, the more annoyingthe echo becomes.

15


16/26

Echo

Echo is normally caused by a mismatch in impedance from thefour-wire network switch conversion to the two-wire local loop.

PSTN is regulated with echo cancellers and a tight control onimpedance mismatches at the common reflection points.

Telephony networks in those parts of the world where analogvoice is primarily used employ echo suppressors, which removeecho by capping the impedance on a circuit, but this is not thebest mechanism to use to remove echo because you cannot useISDN on a line that has an echo suppressor It cuts off the

frequency range that ISDNuses. 16


17/26

Echo

In todays packet-based networks, you can build echo

cancellers into low-bit-rate codecs and operate them on eachDSP. Cisco VoIP, however, does all its echo cancellation on its

DSP.

To understand how echo cancellers work, it is best to firstunderstand where the echo comes from.

Echo cancellers are limited by the total amount of time they

wait for the reflected speech to be received, a phenomenon

known as echo tail. Cisco has configurable echo tails of 16, 24,32, 64, and 128 ms.

It is important to configure the appropriate amount of echocancellation when initially installing VoIP equipment.

If you configure too much echo cancellation, it will take longer for

the echo canceller to converge and eliminate the echo. 17


18/26

Packet Loss

Packet loss in data networks is both common and expected.

When putting critical traffic on data networks, it is important

to control the amount of packet loss in that network.

When putting voice on data networks, it is important to build anetwork that can successfully transport voice in a reliable and

timely manner.

Cisco Systems developed many quality of service (QoS) tools

that enable administrators to classify and manage trafficthrough a data network. If a data network is well

engineered, you can keep packet loss to a minimum.

18


19/26

Packet Loss

Cisco Systems' VoIP implementation enables the voice router

to respond to periodic packet loss. If a voice packet is notreceived when expected (the expected time is variable), it is

assumed to be lost and the last packet received is replayed, asshown below:

Because the packet lost is only 20 ms of speech, the averagelistener does not notice the difference in voice quality.

19


20/26

Dialog-to-Analog Conversion

Almost all of the telephony backbone networks in first-worldcountries today are digital, sometimes multiple D/A conversionsoccur.

Each time a conversion occurs from digital to analog and back,the speech or waveform becomes less "true.

Today's toll networks can handle at least seven D/Aconversions before voice quality is affected, compressedspeech is less robust in the face of these conversions.

It is important to note that D/A conversion must be tightlymanaged in a compressed speech environment. The only wayto manage D/A conversion is to have the network designerdesign VoIP environments with as fewD/A conversions aspossible.

20


21/26

Tandem Encoding

All circuit-switched networks today work on the premise of

switching calls at the data link layer.

The circuit switches are organized in a hierarchical model in

which switches higher in the hierarchy are called tandemswitches.

In the hierarchical model, several layers of tandem circuitswitches can exist, as shown below:

This enables end-to-end connectivity for anyone with a phone, without

the need for a direct connection between every home on the planet. 21


22/26

Tandem Encoding

Voice degradation occurs when you have more than one

compression/decompression cycle for each phone call. Below,I've provided an example of when this scenario might occur:

VoIPTandem Encoding

A drawback to tandem encoding when used with VoIP is that,if a telephony user at branch B wants to call a user at branch C,

two VoIP ports at central site A must be utilized. Also, twocompression/ decompression cycles exist, which means that

voice quality will degrade.

22


23/26

Tandem Encoding

Tandem switches decompress and recompress the voice, the

voice quality can be drastically affected.

It is easy to avoid tandem compression using Cisco Devices: Cisco IOS has other mechanisms that can simplify management of

dial plans and still keep the highest voice quality possible.

Use a Cisco IOS Multimedia Conference Manager.

Use one of Ciscos management applications, such as Cisco Voice

Manager, to assist in configuring and maintaining dial plans on allyour routers.

23


24/26

Tandem Encoding

VoIP WithoutTandem Encoding

A tie-line does not have to be leased from the telephonecompany to complete calls between two PBXs. If a data

network connects the sites, VoIP can ride across that network.

24


25/26

Summary

This chapter brought up many of the issues surrounding VoIP.

Many of these issues, such as compression/decompression ofthe speech frame and propagation delay, are inherent to VoIP,

and you cant do much to minimize these effects on VoIPnetworks.

With careful planning and solid network design, however, you

can control and possibly avoid many problematic issues. Some

of these issues arejitter, overall latency, handling delay,sampling rates, tandem encodings, and dial-plan design.

25


26/26

ThankYou for listening !

Any Questions?

26

Cisco VoIP - Ronak

Documents

Transcript of Cisco VoIP - Ronak