Telecomunicaciones

JOURNAL OF TELECOMMUNICATIONS, VOLUME 2, ISSUE 2, MAY 2010 99

Imprvoing QoS of all-IP Generation of PreWiMax Networks Using Delay-Jitter Model
H. Dahmouni1, H. El Ghazi1, D. Bonacci2, B. Sansò3 and A. Girard4
Abstract—The topic of this paper is the evaluation of QoS parameters in live Pre-Wimax environments. The main contribution is the validation of an analytical delay-jitter behaviormodel. These models can be used in optimization algorithms in order to provide opportunistic and reliable all-IP networks. It allows understanding the impact of the jitter constraints on the throughput and packet loss in wireless systems. However, we show that the real-time QoS requirements of real-time and interactive services can be avoided to a large degree by controlling only the packetdelay-jitter in a fixed and mobile environment. The QoS metrics have been computed from live measurements in a Pre-Wimax realistic environment (Toulouse/Blagnac Airport). Index Terms— Pre-WiMax technology, QoS metrics, Mobility, end-to-end performance, Analytical Jitter Model, ISM band, Measurements.

——————————

——————————

1 INTRODUCTION
he popularity of interactive and real-time applications hashighlighted the limitations of the Internet infrastructure. The different media types exchanged by these applications have significantly different traffic requirements in terms of bandwidth, delay, jitter and reliability, and require different demand service guarantees from the underlying communication network to offer an acceptable performance. Wireless networks have also seen a tremendous growthin their usage and consequent demand for multimedia applications. The 2nd Generation cellular systems like GSM, which offered circuit-switched voice services, are now evolving towards 3rd Generation (3G) and 4th Generation such as LTE and Pre-WiMax systems that can transmit high-speed data, video and multimedia-traffic [11],[12], [17]. Although ATM has become very popular as the backbone andBackhaul of high-bandwidth networks networks because of the QoS guarantees it can offer, it has not been not been widely accepted as a substitute for the IP stacks used on the Internet. However, the telecommunications industry has been migrating from ATM to IP, ensuring that the IP network can support new services and future traffic growth, and allows users take advantage of low-cost bandwidth fromtelecommunications providers. For many future applications, the delay-variation (jitter) remains one of the most important metric of quality of service, and can have a greater impact on the quality than the
• • • • •

T

network latency and packet loss. Even the small amounts of delay variation or jitter introduced by the network can have a significant impact on application performance. With theevolution towards all-IP architecture [11], [13], [21], operators are thinking seriously about dimensioning and optimization tools based on simple and robust delay and jitter models. Usually, the average network delay and throughput have been used as metrics to optimize the network cost and performance in current design tools [3][4]. This is simply due to the absence of a robust and simple formulafor the jitter so that current network planning and design techniques are mostly based on average delay or loss constraints simply because they can often be easily calculated. As a consequence, the effect of jitter on network structure and operation is not well understood. Getting some qualitative understanding of this QoS requirement will be only possible when we have a fast evaluation methodfor jitter. We present here a first experimental validation of the jitter model presented in [1]. We can use it to gain insight into the impact of the jitter on the network performance. In particular we focus on the impact of the jitter on the other QoS metrics such as: the throughput and the packet loss. The objective of the paper is the benchmarking, modeling and analyzing the jitter constraint...