Background Noise Influence on VoIP Traffic Profile

Abstract

Abstract. Modern audio codecs used in VoIP can improve the listening quality by transmitting the main characteristics of the background noise signal during the silence periods. This traffic has been traditionally neglected in the codec mean bit-rate estimation. Nevertheless, when considering an IP environment, the packet overhead increases significantly the required mean transmission bit-rate. Hence, the transmission of the background noise signal can result into either a poor network resource dimensioning in network planning or in the violation of the SLA traffic specifications in a DiffServ scenario. This paper presents a study on the influence of the background noise signal in the mean transmission bit rate required by conversations in IP networks. A new traffic pattern generation model is presented, for which an analytical expression for the mean bit rate is derived. This model is parametrized for the G.729B and the GSM AMR codecs. Experimental results show that this new model significantly enhances the current mean bit rate estimation. The traffic profile of aggregated audio traffic is also addressed, obtaining results which improve the current ON-OFF aggregated traffic models.

An information theoretic framework for predictive channel reservation in VoIP over GPRS

“…Over the years, the duration of activity and inactivity are generally assumed to obey exponential distribution [26][27][28][29][30]. Subsequently, the generation of voice bursts (a collection of packets) is assumed to be Poisson distributed [26][27][28][29][30]. Two different types of speech samples are considered: interactive and business.…”

“…Two different types of speech samples are considered: interactive and business. The interactive speech samples have exponentially distributed 'on' and 'off' period with mean lengths 1026 ms and 1171 ms; respectively [27]. On the other hand, the business speech samples have exponentially distributed 'on' and 'off' period with mean 1490 ms and 1722 ms; respectively [27].…”

“…The interactive speech samples have exponentially distributed 'on' and 'off' period with mean lengths 1026 ms and 1171 ms; respectively [27]. On the other hand, the business speech samples have exponentially distributed 'on' and 'off' period with mean 1490 ms and 1722 ms; respectively [27]. Simulation results are evaluated for up to n t ¼ 400 users with average traffic intensity b ¼ ½0:05; 0:1 erlang.…”