Learn The Main Principles of QoS in the IP Telephony Networks

The level of development of modern network technologies opens up huge opportunities for providers and telecom operators to develop and grow, offering customers all new services at affordable prices. However, technological progress brings not only new opportunities, but also new challenges. One of such problem is providing customers with different levels of service. This is especially true for voice over IP (IP-telephony), when customers require the same quality of communication as PSTN (public telephone network). Therefore, operators need new tools to manage network resources and control the quality of the provided Service.

VoIP traffic is more affected by overload than data traffic. Currently, the Internet does not provide any means to guarantee the quality of service. There is an obvious need to somehow ensure that during periods of overload, real-time traffic will not suffer, or at least receive a higher priority than non-real-time traffic.

All this proves the extremely high relevance of the attempt undertaken in this work to improve the methods of providing QoS quality indicators.

Principles of IP-telephony

IP-telephony is a technology that allows you to use the Internet or any other IP-network to conduct telephone conversations in real time.

In order to organize telephone communications over IP networks, IP-telephony gateways are used. The principle of operation of IP-telephony gateways is as follows: on the one hand, the gateway is connected to telephone lines. On the other hand, the gateway is connected to the IP network. The gateway receives a telephone signal, digitizes it (if it’s an analog one), significantly compresses, splits it into packets and sends it via IP network to the destination using the IP protocol. Packets coming from the IP network are reverse processed. This allows full-duplex communication.

IP-telephony is based on two operations: converting analog speech to digital form, and vice versa, inside the encoding / decoding device and packaging into packets for transmission over the IP network.

Quality of Service Evaluation

For the traffic maintenance at IP telephony, the following main characteristics that determine quality are distinguished.

The first one is the delay time in signal transmission. The following grades of numerical values of delays are provided:

• 1st level – up to 200 ms – the quality of communication is considered excellent. At the same level in the PSTN network, delays are permissible up to 150-200 ms.

• 2nd level – up to 400 ms – the quality of communication is considered good. When compared with the quality of communication in PSTN networks, the difference will be noticeable. However, if the delays are constantly at 400 ms, then this connection is not recommended for important negotiations.

• Level 3 – up to 700 ms – this quality of communication is considered acceptable for conducting unimportant negotiations. Also, this quality of communication is possible when transmitting packets over satellite communications.

The quality of IP-telephony can be attributed to 2-3 levels, and none of the providers can say for sure that it will provide IP-telephony at the 2-nd level, since the delays in the Internet are very unstable. With greater confidence, we can say about the providers of IP-telephony, working on dedicated channels. The quality of communication in such networks corresponds to 1-2 levels. In this case, it is necessary to take into account the delays in encoding / decoding of the voice signal. Thus, the average total delay in the use of IP telephony is usually in the range 150-250 ms.

In addition, there are other factors on which the quality of communication depends; this is the quality of the microphone and the subscribers’ loudspeakers, and the loss of IP packets during transmission through the channel. But even under such conditions, IP-telephony has an advantage over traditional networks: in case of overloading of IP channels and loss of part of the packets, there is special software or hardware that can correct the signal by interpolating neighboring packets, taking into account the characteristics of the speech spectrum.