The paper "QoS in VOIP and Other Networks" is a wonderful example of a report on logic and programming. Quality of service (QoS) is defined as a network capability to offer improved services to a chosen traffic of network over a number of technologies that include IP-routed network, SONET, 802.1 and Ethernet networks, asynchronous Transfer Mode (ATM) and Frame Relay. The prime QoS goal is to offer priority that includes controlled latency and jitter, dedicated bandwidth, and better characteristics loss. Additionally, it is vital to ensure that priority provision for single or extra flows does not result in the failure of other flows.
Technologies of QoS give the rudimentary building blocks, which will be employed for applications of future business in WAN, campus, and in networks of the service providers (Cisco, n.d. a). QoS with Cisco Implementation of QoS in a Simple Network The simple network architecture initiates the three basic pieces for the implementation of QoS. These pieces include QoS marking and identification methods for QoS coordination from one end to another between elements of a network, QoS in a single element of a network, and QoS accounting, management, and policy functions to administer and control end-to-end congestion across a network.
Cisco is implemented by the use of QoS ISO algorithms (Cisco, n.d. a ). QoS fundamentally enables one to offer improved services to particular flows. This is accomplished by either increasing one flow’ s priority or lowering another flow’ s priority. When applying tools for congestion management, one can raise flow priority by serving to queue and queuing in various ways. The queue tool of management employed for the avoidance of congestion increases priority by eliminating flows of lower-priority before those of higher-priority.
On the other hand, shaping and policing offer flow priority by restricting other flows throughput, when connection efficiency tools restrict large flows to indicate small flows preference. Marking and Classification: marking and identification are attained through reservation and classification. According to Cisco (2008) for one to offer preferred services to a particular form of traffic, the traffics must first be recognized. Then the packet needs to be grouped as marked or unmarked. When the identified packet is unmarked, the classification is based on per-hop traffic, a classification that only pertains to the device which is on, and it is not passed to the subsequent router.
This occurs with custom queuing and priority queuing (PQ). If the packets are marked to be used in the entire network, bits of IP precedence can be set. Congestion Management: policing/shaping, link efficiency, queue management, and congestion management tools offer QoS within the element of a single network. Due to data, video, or voice traffic bursty nature, the traffic quantity sometimes exceeds the link’ s speed. At this point, the router is unable to settle on the right technique to use in handling the situation and thus, tools for congestion management are chipped in to handle the situation. Congestion Avoidance: the algorithm of WRED provides for network interfaces congestion avoidance by offering management of buffer and permitting traffic of TPC to throttle back prior to exhaustion of buffers.
This enhances the avoidance of tail drops as well as issues related to global synchronization, therefore, maximizing the performance of TCP-based application and network utilization (Cisco, 2008). Traffic Shaping and Policing: traffic shaping is employed to generate a traffic flow, which limits the flow potential of the full bandwidth.
This is in most cases applied to prevent the problem of overflow. Shaping is the best way for pacing traffic nearer to 384 Kbps to shun remote link overflow. Traffic beyond the rate of configuration is buffered for later transmission to maintain configuration rate. In policing, traffic that runs beyond the rate of configuration is usually discarded rather than being buffered as it takes place in shaping. Link Efficiency: in most cases, links of low-speed present problems for packets of small size.
For instance, when a voice packet is behind a very big packet, the voice delay budget would be surpassed long before the packet leaves the router. In such cases, link interleaves and fragmentation permit the big packets to be fragmented into minute packets interleaving the packet of voice. Link efficiency can also be assured by the elimination of excess overhead bits. This can be done by the use of RTP compression of the header that lowers the header to a manageable size (Cisco, 2001). Signaling: Cisco provides QoS management software for the IntServ model of IETF.
This model contains RSVP, which is a basic mechanism to execute flows admission control in a network. According to Cisco (2008), this technology is best suited in VOIP, where it controls the availability of call resources. According to VOIP configuration, a call can only be completed if all resources required are available, this technique ensures any call getting into the system does not collide with or influence existing calls ‘ quality. QoS policy proliferation through BPP is another technique that permits direct signaling the forwarding priority for IP-prefix, AS-path, or autonomous destined packets by use of the BGP community list aspect. Traffic Conditioning: traffic getting into a network can be controlled by the use of a shaper or policer.
A shaper restricts the flow of traffic while a policer imposes a rate-limit to its particular rate by use of buffers. FRTS, GTS, and CAR mechanism can be configured within/without the framework of QMC. Link Competence Mechanisms: voice traffic and steaming video employ the RTP. RTP, UDP, and IP packet headers may be compressed from about 40 down to 8-5 bytes.
This preserves a remarkable quantity of bandwidth in the low-speed connections cases, and while a big number of streams of media are being supported. Additionally, FRF. 12 frame fragmentation specification and LFI specification permit fragmentation of big packets of data, interleaving them with packets of RTP, and maintaining low media streams jitter and delay (Cisco, 2006a).