8.0 Conclusions
The telecom and datacom industries are undergoing massive transformations, occasioned by the emergence of communication ecosystems that have evolved from previously isolated and application specific networks. Telecommunications has since ceased to a utility service, and instead evolved into converged video, voice and data networks. The demand for video and voice traffics over the internet has placed strains over the available resources, not only because of their relatively large size and demand volume, but perhaps even most significantly, because of the need to minimize packet delays or losses in order to facilitate interactivity and real time communication needs. The growth in multimedia traffic presents obvious technical and practical problems by way of streaming, reliability, end-to-end latency, integrated services and multipoint communication requirements that are critical for the viability of video/audio-based communication over the internet. These technical problems are brought on by even more subtle, and growing expectations among the consumers of internet services, who need more speed, quality, efficiency, availability etc. Quality of Service for some or all multimedia traffic is central to meeting the increasing challenge of a new landscape in communication, and better management of the existent network resources.
In order to ensure QoS, it is evident according Zamora, Jacobs, Eleftheriadis, Chang, & Anastassiou (2000); (DeCusatis & Jacobowitz, 2006) and (Ergin, Gruteser, Luo, Raychaudhuri, & Liu, 2008) and the wider material presented in this paper, service differentiation is indispensable, and it is the building block for increased efficiency in the utilization/management of the existent and future network resources. It ensures moderated delays, data packet losses or even abortion of entire transmission operation in the interest of averting traffic competition or QoS violations to the detriment of priority traffic. Converged networks necessitate efficient allocation of bandwidth and other network resources to multiple service needs, that in turn require one or more QoS mechanisms.
Save for the electronic switching limitations, costs and poor scalability, over provisioning is the most straight forward approach to avoiding congestion, especially for LANs, but for larger, more intricate connections, efficient resource management, routing and traffic engineering are the most effective mechanisms, are even made better through combinations (Ergin, Gruteser, Luo, Raychaudhuri, & Liu, 2008). In this respect, Best effort is over provisionings equivalent, simple and equally efficient on a limited scale. DiffServ offers greater scalability and it is a per-aggregate-class service discrimination technique employs packet tagging compared to IntServ. DiffServ also benefits from the flexibility stemming from the separation of policy and supporting mechanisms. Traffic policing makes is crucial in avoiding video and other traffic QoS violations, especially non-real time traffic, which has not quantifiable traffic parameters and requires as much bandwidth as possible to the detriment of video/audio/real time traffic, Zhao, Olshefski, & Schulzrinne (2009). This can similarly be achieved through admission control, which disallows QoS violations, by smoothing traffic flows, despite the fact that perform poorly in the case of very unpredictable and widely varied data. Bandwidth brokerage, QoS routing, scheduling, queuing management, multi-protocol label switching and label distribution controls are equally vauable QoS mechanisms that have been, or can be deployed to assurance. quality and reliable video traffic over the internet, without affecting the efficiency with which other forms of data are trasmitted. The practical utility and potential of MPLS is massive owing to its ability to offer the best of the functionality that is available from an overlay model in a manner that is integrated, without significant cost implications. This is also evidenced in Jaffar, Hashim, & Hamzah (2009); Dar & Latif (2010); Stiller (2009) and Jaffar, Hashim, & Hamzah (2009), which shows that DiffServ-MPLS results in mutiply greater efficicncy for video traffic, including quality, efficincy, scalability and potential among other performance variables.
The tradeoff between scalability and QoS, QoS and costs and well as QoS and practicality are evident through this paper, and they are important consideration in the choice of the technologies to be used for ensuring the quality of service for video and other traffic, if at all. In addition, the ability and application of these mechanisms are technologies have been widely tested. Jaffar, Hashim, & Hamzah (2009) for instance studied their application in coverged networks, determining that they were fairly useful. QoS is of particular important in mobile phones, where IXIA (2011) proved that the mechanisms were practably efficient. The security issues to video and other applications have been raised especilly in in the PSTN technologies and VoIP. It is clear that using security protocols to secure traffic on nentworks slows down traffic, which effectively places an additional strain on the available resources, and the potential of some QoS mechanisms to be used without scaling up network infrastructure, Fiandrotti, Gallucci, Masala, & Magli (2009).
While networking resources are scarce, there is sufficient capacity to facilitate quality video and audio traffic transmission, but it is clear that the biggest challenges that face the emergence of multimedia traffic remain security and quality. There are varied avenues of research that need to explored, which may have great results for the industry in the future. These include the management and arhitecture for differentiated traffic engineering (DTE) that could have massive implications on the video quality of service provisioning. Development of DTE monitoring and management algorithms and rules rules for instance, to used in the request for modified paths presents a reach research area that could greatly expedite the developments in video QoS. Other areas that require further research include DiffSer-MPLS, wider MPLS and other similar technologies, which according to Jaffar, Hashim, & Hamzah (2009), need to be explored further.
