Other Free Encyclopedias » Online Encyclopedia » Encyclopedia - Featured Articles » Contributed Topics from P-T » Quality of Service in Multimedia Networks - QoS Elements, Realization of QoS, Network Scalability and Application Layer Multicasting

Network support and protocols for QoS

service diffserv layer forwarding

Recently, there has been much research conducted on the development of IP QoS (Quality of Service), keeping in mind the goal of allowing network operators to offer diverse levels of treatment to multimedia packets. These efforts have caused several approaches to emerge: the Integrated Services (Intserv) and its accompanying signaling Resource Reservation Protocol (RSVP), the Differentiated Services (Diffserv) desired to accommodate heterogeneous application requirements and user expectations, and the permission of differentiated pricing of Internet services, and Multi Protocol Label Switching (MPLS).

Intserv is basically an extension of the original Internet architecture to support new types of real-time applications, and to improve the network manager’s ability to control the network. The IntServ model enables applications to choose from multiple, controlled levels of delivery service for their data packet. Intserv model includes the Control Load (CL) service, Guaranteed (G) service, and Best Effort (BE) service to define QoS along the network. TheGuaranteed service is defined for applications with rigid end-to-end constraints, and ensures that packets arrive within the guaranteed delivery time and that they will not be discarded dueto buffer overflows. The Controlled-Load service, on the other hand, is defined for applications with looser performance criteria, and offers a commitment to a service similar tothat provided to best-effort traffic under lightly loaded conditions. IntServ works in conjunction with a signaling protocol such as Resource Reservation Protocol (RSVP). In thisarchitecture, the sender and the receiver exchange data in order to negotiate specific packet classification and forwarding behavior in the network. Once a path is established, the sendertransmits the data knowing that resources have been reserved for its packets all the way to the receiver. The routers in the network need to store state information for every sender-receiverpath passing through them, which renders this scheme less scalable in comparison to others. The problems with RSVP architecture is that it does not scale well in the Internet core, sincethe amount of information increases proportionally with the number of flows, which imposes huge storage requirements and processing overhead on the routers.

DiffServ is an architecture that provides service differentiation. Using DiffServ, applicationscan mark the first six bits of the type of service (TOS) field in the IP header for requesting a higher level of service. These marked packets are then forwarded through the network bymapping the 6-bit TOS value to a forwarding behavior supported in a Domain router. In this model, the role and importance of edge routers and core routers are more clearly definedcompared to the aforementioned approaches. The Diffserv model introduces a mechanism to aggregate micro-flows into a number of predefined set of building blocks from which avariety of aggregate router behaviors may be designed to provide quality of service. IP packets are tagged with distinct labels before entering an IP Diffserv domain and given aparticular forwarding treatment at each network node along the path. Currently, the Diffserv model defines three standard Per-Hop-Behaviors (PHBs): the Expedited Forwarding (EF), the Assured Forwarding (AF), and the Best-Effort PHB. In practice, there existtwo models for implementing Diffserv on networks: the relative Diffserv, and the absolute Diffserv model.

Using Label-switching technique , a forwarding ‘label’ (link-layer addressing) is attached to the packet that contains the next hop address in the network. This label also determines the forwarding behavior of the packet and the value of the replacement label at the next hop. An example of this approach is MPLS (Multi-protocol Label Switching), which is a technology that integrates the label-swapping paradigm with network-layer routing. Using MPLS is another way of receiving service differentiation. Multiprotocol implies that MPLS supports various protocols at both, Layer-2 (link layer) and Layer-3 (network layer) of the OSI (Open Systems Interconnection) model. Layer 2 protocols supported include ATM, Frame Relay, etc. and some of the Layer-3 protocols supported are IPv4, IPv6 and IPX.


User Comments

Your email address will be altered so spam harvesting bots can't read it easily.
Hide my email completely instead?

Cancel or