This subsection displays the results of the fourth applications’ quality scenario, where 3D technology was implemented with the video contents and the web browsing applications. Figure 4.30 shows that this move to 3D technology has significant influence on the throughputs of the different applications. Here, the streaming video has the largest share of this influence, where its consumption of bandwidth is increasing more and more by this quality improvement. Again this largest influence of streaming video by the quality improvement is attributed to the same two main reasons, the popularity of this applications type and the full inter-node traffics direction.
Figure 4.30 Total Network's Load Rate due to the Different Application Types
Even though the web browsing applications have been upgraded to 3D, the rise in the throughput generated by web browsing is still incomparable with the rise of the throughput generated by steaming video. This is due to small increase in the per user bit rate when web browsing is switched to 3D which is met by the large increase in the per user bit rate of streaming video when it is moved to 3D (see Table 4.5).
No wonder, this high bandwidth streaming video traffic will be more concentrated on the link [Server-Basra] as it is the only gateway from the server causing this link to be highly congested at the busy hour compared to the others as shown in Figure 4.31.
Figure 4.31 Total Load Rate on the Different Network's Links at Busy Hour
Figures 4.32 and 4.33 show the busy hour load rates on the link [Server-Basra] at the busy hour and in per hour bases during the whole day respectively. It is obvious in both figures that the streaming video traffic dominates and governs the total rate value and per hour variation shape.
Figure 4.32 Busy Hour Load Rates on the Link Server-Basra due to the Different Application Types
Figure 4.33 Per Hour Load Rate on Link Server-Basra
In the link [Basra-Ashar], the existence of the non-server-based busy hour traffics did not add substantial contributions to the total load since the streaming video’s load is much higher than any other traffic type in this scenario as shown in Figure 4.34.
Figure 4.34 Busy Hour Load Rates on the Link Basra-Ashar due to the Different Application Types
The significance of the traffic generated by all the applications other than the streaming video has been lessened in this scenario due to the huge gap between the throughputs of streaming video and all the other applications. However, the low usage of streaming video service confronted by the high level of the miscellaneous video clips up/downloading usage at the early day’s hours made the influence of the latter to appear on the per hour total load rate on this link as shown in Figure 4.35.
The link [Basra-AlZubair] also holds traffics of all types. Again, this link is similar to the link [Basra-Ashar] in terms of the dominance of streaming video load rate, but the lesser destination’s population has led to lower bandwidth consumption for all the applications as shown in Figure 4.36.
Figure 4.36 Load Rates on the Link Basra-AlZubair due to the Different Application Types
Figure 4.37 shows the hourly variation of load rates due to the different application types together with resultant total per hour throughput on the link [Basra-AlZubair]. This figure indicates that the total per hour rate shape is governed by the dominant streaming video application’s traffic after 8:00 AM. On the other hand, it is more dominated by the miscellaneous video clips up/downloads’ traffic before 8:00 AM because of the increasing use of this application in this period of time as shown in Figure 4.37.
Figure 4.37 Per Hour Load Rate on Link Basra-AlZubair
4.10 Summary
Basra national optical network has been simulated in this chapter. The total average load rates on the different network’s links have been calculated, and the bandwidth of the traffics generated by the different applications types has also been determined against different application’s quality input scenarios. Furthermore, the contribution of each type of applications to the total network load rate has been dimensioned for each resolution scenario. The links’ peak-hour bit rates obtained for each of the four scenarios in Figures 4.3, 4.13, 4.23, and 4.31 provide a clear indication of the required dimensioning of these transmission links.
The links’ throughputs due to the different application types have been calculated both at busy hour and on an hourly basis during a 24 hour period, and the total link’s throughput through the whole day has also been investigated accordingly. In addition, the places of potential congestions on the different network’s links together with the contribution of each application type to that congestion have also been revealed in this chapter.
The effect of enhancing the quality of the services offered to users on the network loading has been studied extensively aiming to put a corner stone for establishing an efficient network that satisfies the demands of future internet which is expected to be characterised by its huge bandwidth consumption yields due to the enhancements in the offered services qualities and the dramatic increase in the number of users.
Chapter 5
The Impact of Content Oriented
Routing on OpenFlow Burst Switched
Optical Networks
The Internet has revolutionized how the world has accessed information and communicated it but continued development in optical and electronic technologies has resulted in higher bit rates, bandwidths, and functionalities and has enabled the spawning of a multitude of different applications and services that can be accessed on it. This has exposed fundamental shortcomings to the existing Internet architecture.
Three shortcomings of the current Best Effort (BE) routing protocol, which is widely used on Internet currently, has been addressed in this chapter. First, the equal treatment of all traffic types can be considered a great deficiency in the BE protocol since applications sensitivities towards delay and packet loss are different [84]. This chapter proposes dealing in different ways with the different traffic types to offer an enhanced QoS. Secondly, the blind view of BE routing protocol to the type of traffic results in an unbalanced distribution of traffic over network links causing some links to become congested and others not and is a second drawback in the current Internet [85]. Weighted Dijktra Algorithm was deployed in this chapter to balance the load over the links, which in turn will decrease the amount of loss and thus enhance the QoS. Finally, another challenge was tackled through this chapter to provide an extra free space over the network's links to widen the scope of services being offered to the users, such as including broadcast multicasting, whilst maintaining the QoS within an acceptable range. A lot of Internet
providers have struggled to include TV broadcasting with their services, since most of these attempts failed to provide a known QoS [86]. Balancing loads over links will also help to increase link utilisation in the network. Finally, the propagation delays and traffic losses for traffics over the links were recorded to prove the feasibility of our new traffic management technique, represented by the proposed content oriented routing, for the future Internet. Loads over links were also monitored in this work to find out to what extent our load balancing technique helps to relieve network's congestions.
Section 5.1 introduces the idea of application based routing. Section 5.2 is concerned with network modelling and how the real national network had been implemented in the software. Furthermore, this section lists the types of applications used in our model and categorizes these applications into three main classes according to their QoS requirements. In Section 5.3 the network’s performance was analysed and the results were discussed. Finally the Section 5.4 concludes this chapter.
-What’s new: A routing protocol was suggested to maintain the different applications in
our flow model within pre-set QoS levels.