Premium
Predictive model for hardware calibration to transmit real‐time applications in VoIP networks
Author(s) -
Gupta Nishant,
Mahajan Nitish,
Kaushal Sakshi,
Kumar Naresh,
Kumar Harish,
Sangaiah Arun Kumar
Publication year - 2019
Publication title -
concurrency and computation: practice and experience
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.309
H-Index - 67
eISSN - 1532-0634
pISSN - 1532-0626
DOI - 10.1002/cpe.5190
Subject(s) - voice over ip , computer science , quality of service , codec , computer network , packet loss , network packet , jitter , real time computing , the internet , computer hardware , telecommunications , world wide web
Summary Voice over Internet Protocol (VoIP) carries and transforms voice over the IP networks. The principles of VoIP calls are similar to traditional telephony that involves signalling, channel‐setup, digitization, and encoding of speech signal, but it transmits data over a packet‐switched network instead of circuit‐switched network. Factors which determine VoIP Quality of Service (QoS) include the choice of codec, packet loss, delay, jitter, and optimal hardware selection to handle different services. Hardware Calibration is a mechanism used for selecting an appropriate hardware for call manager to process and transmit different applications in real time. The widespread use of VoIP services in formal and informal sector produces a significant amount of data with variety of dimensions. This data can be used as leverage to analyze the system and predict various factors boosting the performance and cost effectiveness of the system. This paper proposes a predictive model that selects the best suitable hardware to handle particular offered load, which can support desired numbers of concurrent calls from wide array of processors available in the market today. This model would help the VoIP service providers in providing efficient services with QoS for different VoIP services like voice, data, video, chat, etc. This paper attempts to train and evaluate a model using various system parameters and system benchmark is predicted on an absolute scale. The results effectively demonstrate the selection of best call manager to handle offered load and hence provides QoS in overall network performance.