Traffic Management Of IPTV System

In several recent years, more and more applications and communications are requiredto responsible for greater convenience of the people life such as real-time applications,interactive applications, mail, web browser, file transfer… They are characterized by itscapability of providing many different levels of Quality of Service (QoS) for the users thatcould choose what they would like to be provided and how much they want to pay fortheir services. Supporting QoS in Internet nowadays is not really a simple task because ofstrict disciplines for real-time applications as jitter, end to end delay, packet loss…This is done by implementing some QoS techniques in the heterogeneous networkwhich promise low latency will be achieved for each application requirement. And themost popular method is known as Differentiated Services (DiffServ). DiffServ architectureis composed of two key components: traffic conditioning and Per Hop Behavior (PHB).Traffic conditioning is responsible for several tasks when the router receive packets frominput interface and treat them step by step as follows: classifier, meter, marker, shaper anddropper. PHB include some queuing mechanisms that service respectively with thepriority of each application such as: First In First Out (FIFO), Priority Queuing (PQ),Weighted Fair Queuing (WFQ), Weighted Round Robin.This thesis addresses the impact of several mechanisms in DiffServ into trafficmanagement for IPTV system. Solving the congestion take place in access network andoptimizing network resource utilization are two main tasks of this context. OPNETsimulator is one of the most popular tools to simulate network. Various schedulingmechanism and queuing management algorithms are experimented under differentnetwork traffic scenarios in order to evaluate time constraint of real-time applications suchas voice and video while providing simultaneously best effort services. Furthermore,simulation results will be compared their relative performances base on queuing delay,probability of packet loss, link utilization… Base on analysis results, managementmechanism can be chosen suitable for each traffic profile in IPTV network.