Tamal Chakraborty

Study of enhanced VoIP performance under congested wireless network scenarios

Stringent QoS maintenance in a wireless environment for VoIP communication is a major challenge. VoIP inherently generates constant bit rate traffic and is highly sensitive to network delay. However, unpredictable network congestion makes a VoIP session so degraded that its QoS goes below a tolerable limit. Accordingly, a suitable solution is needed to adapt varying network conditions satisfying minimum QoS while maintaining transparency to the end user. We have extensively studied the effect of variable voice packet payload size with changing number of voice sample frames in the payload of the RTP packets. For this, we have used OPNET Modeler 14.5.A to observe the performance in terms of MOS, End-to-end delay through extensive simulations. Results are provided with valid discussions. Based on the observation, we propose an Adaptive VoIP (AdVoIP) algorithm that may be used in any real life VoIP network for further enhancement of the performance. The adaptive algorithm may use RTCP Receiver Reports to assess the network conditions in real wireless scenarios.

Optimization technique for configuring IEEE 802.11b access point parameters to improve VoIP performance

The performance of wireless LANs is greatly affected by path loss, RF interference and other sources of signal attenuation in addition to network congestion. The primary factors involved in effective real-time communication, namely delay and loss, must be within certain controlled limits in such a scenario. In this paper, we analyze the various factors driving IEEE 802.11b access points through extensive simulations and thereafter develop an optimization technique to configure the parameters of the Access Point. We simulate our test bed scenario and apply the developed algorithm. Finally, we implement the configured parameters in our testbed to provide optimum Voice over IP (VoIP) performance. Simulation and measured results have been included.

Design and study of VoIP model in cognitive radio network under different simulation platforms

Rapid growth followed by increasing popularity in VoIP systems has witnessed rollout to new services and applications in emerging technologies like Cognitive Radio Networks (CRN). The stringent QoS requirements in VoIP coupled with the complexities in CRN have initiated intensive research in the field of performance analysis and optimizations guided by simulation results. However, in the absence of any standard model of VoIP over CRN, accuracy and credibility of the simulation output are strongly dependent on proper design of the simulation model. The objective of this paper is to build standard models for VoIP in CRN and successfully implement VoIP applications over CRN domain, which will serve as initial point for development with respect to all future simulation studies in VoIP over CRN category. Initially, models of VoIP in CRN are developed using OPNET Modeler 16.0.A following distributed architecture in single-channel and multi-channel scenarios. The work is extended by designing VoIP over CRN model in Visual C++ adhering to the principles of centralized architecture. The models are validated by comparison of simulation results obtained in both platforms. Extensive analysis of simulation results in the developed models indicates significant trends in VoIP QoS metrics with variation in CRN parameters.

Analysis and enhancement of QoS in cognitive radio network for efficient VoIP performance

VoIP service demands strong QoS support for enhanced call quality and increased user satisfaction. Deploying VoIP in opportunistic communication models like cognitive radio requires thorough analysis of the factors involved in design and implementation. The objective of this paper is to implement VoIP successfully over a cognitive radio network and maintain the call quality at an acceptable limit. Initially, a model of VoIP is built over cognitive radio network using OPNET Modeler 16.0.A. The QoS parameters of delay, jitter and packet loss are analyzed over a basic cognitive radio cycle in the developed model. Modifications are proposed in the cognitive radio cycle for improvement in call quality and gain in throughput. Extensive simulation studies reflect overall QoS enhancement after application of the modified cycle. Finally, the proposed approach is further enhanced by modifying the VoIP parameters that result in high quality VoIP calls as witnessed from the simulation results.

Design and analysis of channel reservation scheme in Cognitive Radio Networks

The primary aim of a Cognitive Radio Network (CRN) is to provide higher spectrum utilization through an opportunistic usage of the licensed spectrum bands by the unlicensed Secondary Users (SUs). However, these SUs face unpredictable interruptions in their transmissions due to sudden arrival of the licensed Primary Users (PUs), resulting in either handoff or dropping instances. The channel reservation policy for the PUs is one such solution that reduces the probability of their interference, though at the cost of lower system throughput. This paper provides a complete study of the channel reservation scheme in a CRN. Initially, the system model is defined along with an architectural framework to address the practical constraints involved in implementing the channel reservation strategy. A novel design approach is then adopted to develop the mathematical models, followed by an evaluation of the network parameters to obtain a tradeoff between the decrease in interference and low spectrum utilization when some channels are reserved in the CRN. The optimal number of channels to be reserved is also determined. In addition, as a potential application of the channel reservation policy, an analytical framework is designed for the VoIP applications in the CRN, that highlights significant improvement in the VoIP call quality. Finally, an extensive analysis in the real-life like simulation models validates the inferences drawn from the mathematical models, recording over 50% improvement in the interference-free transmission of the SUs with the channel reservation scheme in the CRN.

Proactive QoS Enhancement Technique for Efficient VoIP Performance over Wireless LAN and Cognitive Radio Network

VoIP over Wireless LANs is greatly affected bypath-loss, RF interference and other sources of signalattenuation in addition to network congestion. The primaryfactors involved in effective real-time communication,namely delay and loss, must be within certain controlledlimits in such a scenario. Wireless LAN Access Points must,therefore, act in tune with the codecs to ensure high qualityof the ongoing VoIP sessions. Further deployment of VoIPin opportunistic communication medium like cognitiveradio network requires thorough analysis of the factorsinvolved in design and implementation of such networksalong with modifications in VoIP parameters. In this paper,the various codec parameters are analyzed with respect towireless LAN access points. Thereafter, an optimizationalgorithm is developed based on proactive strategy to keeploss and latency within tolerable limits. The proposedalgorithm is hence implemented in the SIP enabled test-bedto provide optimum Voice over IP (VoIP) performance. Theproactive strategy is further applied to study VoIPperformance over cognitive radio networks. A simulationmodel is developed for VoIP transmission in cognitive radionetworks with two scenarios where the primary and thesecondary user involve in VoIP transmission respectively.For each scenario, extensive analysis of both VoIP and CRNparameters is performed and an algorithm is proposedbased on codecs and active queues to ensure highthroughput of VoIP traffic with enhanced QoS. Resultsverify significant performance improvement in VoIP callquality under both wireless LAN and cognitive radionetwork environment. Simulation and measured resultshave also been included.

Designing a real-time spectrum handoff algorithm for VoIP based Cognitive Radio Networks

Secondary Users (SUs) in a Cognitive Radio Network (CRN) face unpredictable interruptions in transmission due to the random arrival of Primary Users (PUs), leading to spectrum handoff or dropping instances. An efficient spectrum handoff algorithm, thus, becomes one of the indispensable components in CRN, especially for real-time communication like Voice over IP (VoIP). In this regard, this paper investigates the effects of spectrum handoff on the Quality of Service (QoS) for VoIP traffic in CRN, and proposes a real-time spectrum handoff algorithm in two phases. The first phase (VAST-VoIP based Adaptive Sensing and Transmission) adaptively varies the channel sensing and transmission durations to perform intelligent dropping decisions. The second phase (ProReact-Proactive and Reactive Handoff) deploys efficient channel selection mechanisms during spectrum handoff for resuming communication. Extensive performance analysis in analytical and simulation models confirms a decrease in spectrum handoff delay for VoIP SUs by more than 40% and 60%, compared to existing proactive and reactive algorithms, respectively and ensures a minimum 10% reduction in call-dropping probability with respect to the previous works in this domain. The effective SU transmission duration is also maximized under the proposed algorithm, thereby making it suitable for successful VoIP communication.

A Priority Based Adaptive Channel Reservation Algorithm for Improved System Capacity in Cognitive Radio Networks

Cognitive Radio Network (CRN) aims to solve the problem of spectrum congestion by allowing opportunistic use of idle spectrum bands by secondary users (SUs) in the absence of primary users (PUs). Channel reservation policies have been proposed in this regard to ensure maximum throughput for these SUs, while minimizing their interference with PUs. However, these algorithms always reserve a fixed number of channels for PU, leading to under-utilization of idle channels in times of low PU activity. This paper addresses this problem by introducing a novel Priority based Adaptive Channel Reservation (PACR) algorithm that adaptively reserves channels based on PU activity and further performs priority based channel allocation to both real-time and non real-time users. Mathematical models are designed for PACR based CRN, that depict increase in system capacity and system heterogeneity with respect to hosting both Voice over IP (VoIP) and data applications. Extensive performance analysis obtains trade-off among critical system parameters and ensures maximum Quality of Service for the VoIP users. Finally, analysis in real-life like simulation models validates the inferences drawn from the mathematical models, recording over 100% improvement in SU throughput for PACR algorithm as compared to static channel reservation policy.

An analytical framework for channel reservation scheme in Cognitive Radio Network

The primary aim of Cognitive Radio Network (CRN) is to provide higher spectrum utilization through opportunistic usage of licensed spectrum bands by unlicensed (secondary) users. However, secondary users face unpredictable interruptions in transmission due to sudden arrival of primary users, resulting in either handoff or dropping instances. Channel reservation for primary users is one solution that reduces the probability of their interference with secondary users, though at the cost of reduced system throughput. This paper addresses the tradeoff between decrease in interference and low spectrum utilization in channel reservation scheme by designing mathematical models to implement channel reservation strategy in CRN. Comparative performance evaluation of the developed models is performed to obtain optimal tradeoff between system parameters. The models are further analyzed for optimal number of channels to be reserved for primary users. Practical constraints involved in implementing channel reservation scheme are discussed in detail. Performance analysis in real-life like simulation model validates the inferences drawn from the mathematical model.

Simulation Design and Performance Analysis for VoIP in Cognitive Radio Networks

Increasing popularity of VoIP systems has witnessed applications in emerging technologies like Cognitive Radio Networks (CRN). The stringent QoS requirements in VoIP coupled with complexities in CRN have initiated intensive research in the field of performance analysis and optimizations guided by simulation results. However, in the absence of any standard model of VoIP over CRN, accuracy and credibility of the simulation output are strongly dependent on proper design of the simulation model that must have a strong mathematical foundation. The objective of this paper is to build standard models for VoIP in CRN and successfully implement VoIP applications over CRN domain, which will serve as initial point for development with respect to all future simulation studies in VoIP over CRN category. Initially, models of VoIP in CRN are developed using OPNET Modeler 16.0.A following distributed architecture in single-channel and multi-channel scenarios and further in Visual C++ adhering to the principles of centralized architecture. The models are validated by comparison of simulation results obtained in both platforms. The underlying mathematical model behind the design is established and the critical factors pertaining to both VoIP and CRN domain are extensively analyzed.