Printer Friendly

ANALYTICAL MODEL OF A BURST ASSEMBLY ALGORITHM FOR THE VBR TRAFFIC IN THE OBS NETWORKS.

Byline: M.A.Al-Shargabi, A.Abid and H.Mellah

ABSTRACT: This paper presents a proposed analytical model for the number of bursts aggregated in a period of time in OBS networks. The model considers the case of VBR traffic with two different sending rates, which are SCR and PCR. The model is validated using extensive simulations, where results from simulations are in total agreement with the results obtained by the proposed model.

1. INTRODUCTION

All-optical networks seem to be rational choice for the Internet backbone infrastructure because of their characteristics, such as high transmission speed, and huge bandwidth. Dense Wavelength Division Multiplexing (DWDM) is one of the optical networks technologies, which offers a high transmission speed and huge bandwidth. The optical packets in DWDM can be switched with one of three optical switching methods that are Optical Circuit Switching (OCS) [1], Optical Packet Switching (OPS)[2], and Optical Burst Switching (OBS) [3-4].

OBS is a switching method that combines the benefits of the circuit switching and packet switching methods [3]. In OBS network, Incoming traffics (data) from clients are aggregated to create data burst at the edge router of the OBS source network (ingress) using the burst assembly process. The burst header (control) transmitted slightly ahead in time in order to allow optical core nodes to process the headers electronically for the establishment of an end-to-end optical path, and to switch data bursts optically. After an offset time, the ingress node sends the data in the form of bursts. OBS uses one way signaling scheme with out-of-band method. The OBS data bursts may have variable data lengths, and handle several types of traffic (IP packets, ATM cells, Frame Relay frames, etc.) with their distinguished characteristics and constraints.

An important design aspect of an OBS network is the burst assembly process performed at the edge nodes. This process concentrates on the upper layer packets which are optically transmitted over the OBS network. In view of this, the burst assembly is a process responsible for gathering the packets to create the data burst inside the OBS network. The ingress node performs the burst assembly by arranging the packets which arrived at different destination queues. Using different techniques, the ingress node sends the data burst to the core nodes (switch) as one unit. Thus, fewer requests arrived at the core nodes, causing a reduction in the signalling overhead. As a result, slower and cheaper switching technologies can be applied.

Moreover, the OBS burst assembly leads to control the traffic characteristics such as the burst interval time and length which will affect the network performance. Furthermore, the burst assembly provides the ability to integrate the characteristics of the traffic type in the burst assembly process. Different strategies have been proposed to perform the burst assembly process. The majority of these strategies are time-based, threshold-based or both time- and threshold-based.

The time-based strategy uses an interval time (T) to create the data burst. All packets which arrived in the interval time, T will be gathered to create the burst. Each destination queue has its own timer which is set to start from 0. During the interval time, all the packets with the same destination will be arranged based on their destination. When the timer reaches T, all the packets in the queue will be incorporated into the data burst. However, the drawback of this strategy is the long burst length. If the input traffic is high, it increases the loss rate in the core node. In the low input traffic, the small burst size will create a large number of control packets which increase the process in the core node.

In contrast, the threshold-based strategy uses the maximum size Bmax parameter to create the data burst. All the packets will be arranged in the destination queue until the number of bytes reaches the Bmax parameter value. The weakness of this strategy is the delay time which the burst takes to reach the Bmax parameter value.

Both the time- and threshold-based strategies are proposed to overcome the limitation of the previous two strategies. In the time- and threshold-based strategies, the burst is created either when the timer reaches the time T value, or when the number of bytes reaches the Bmax value. Nevertheless, if the timer reaches the time T value in the first case, it should compare the number of bytes with the third parameter Bmin which represents the minimum number of bytes to create the burst. The number of bytes should be greater than Bmin as compared to all the packets in the destination queue gathered to create the data burst. Otherwise, the creation of the data burst is postponed until the next interval time.

One of the main advantages of the OBS is the ability to support different types of traffic with distinguished characteristics and constraints. The VBR traffic is one of the Internet traffic that can be carried in the OBS network, rendering a solution to reduce the number of transmission and storage. In the VBR traffic, many issues must be taken into account when configuring the network such as traffic contract. The traffic contract in the VBR traffic can be applied by specifying some values such as the Peak Cell Rate (PCR), Sustainable Cell Rate (SCR), and Maximum Cell Transfer Delay (MaxCTD). Applying these parameters could affect the assembly process.

Various studies [5-8] have given a great attention for the burst assembly process and the effect of the burst assembly algorithms on the characteristics of the traffic.

The rest of the paper is organized as follows. Section 2 introduces the Variable Bit Rate (VBR) traffic. Section 3 introduces the proposed analytical model. Section 4 introduces the results and discussions.

Finally, section 5 concludes the paper.

2. VARIABLE BIT RATE (VBR) TRAFFIC

Video and audio communications like video conferencing, news, and chat applications will occupy greater portions of the Internet traffic. Thus, using compressed video and audio traffic seems to be compulsory to reduce the costs of transmission and storage.

VBR compressed video and audio traffic have been receiving considerable attention due to their compression algorithms that can transmit with high rate during the high activity and with low rate during the less activity, as shown in Fig.1 .

VBR has been known as an ATM service category (Traffic Management Specification, 1996) intended for both the real- time and non-real-time applications. The Real Time-VBR (RT-VBR) service category is used for the connections that transport traffic at variable rates and rely on accurate timing between the traffic source and destination.

Real-time VBR connections can be characterized by Peak Cell Rate (PCR), Sustainable Cell Rate (SCR), and Maximum Burst Size (MBS). The delayed packets can be specified by Maximum Cell Transfer Delay (MaxCTD) parameter.

The Non Real Time -VBR (NRT-VBR) service category is used for the connections that transport traffic at variable bit rate and there is no inherent reliance on time synchronization between the traffic source and destination, but there is a need for an attempt to guarantee a specified bandwidth or latency. The VBR traffic grants some advantages as compared to the constant bit rate (CBR) traffic, as it offers higher quality and greater chance for statistical multiplexing [10-12].

3. VBR Traffic Parameters

VBR traffic can be configured using four traffic parameters PCR, SCR, MaxCTD, and MBS [13-15]. The PCR and the SCR are the upper limit and the average of bits of data on the source traffic that can be sent, respectively. The MaxCTD is a parameter that determines the packet maximum delay in the network.

Whereas, the MBS defines the amount of time or the duration at which the source sends at the PCR. VBR Traffic Forms VBR traffic can be formed in three different forms that are high traffic rate, sustainable traffic rate, and burst traffic rate. The sustainable traffic rate form is used when there are fewer activities, so the source transmits the data at the rate or near the rate of the SCR as shown in fig. 2.a. On the other hand, the high traffic rate form is used when there are high activities, so the source transmits the data at the rate or near the rate of the PCR as shown in fig. 2.b. In contrast, the burst traffic rate form is used when there are continual changes in the activities, so the source transmits the data at the rate between the SCR and PCR as illustrated in fig. 2.c.

4. THE PROPOSED ANALYTICAL MODEL

An analytical model to obtain the distribution of the number of bursts created by the assembly algorithm using both time- and threshold-based strategies was proposed in [16]. The model assumes that the packet arrival process is Poisson. The resulting burst process is characterized as a bulk burst transmission that is slotted in time.

The Real Time-VBR (RT-VBR) service category is used for the connections that transport traffic at variable rates and rely on accurate timing between the traffic source and destination.

Real-time VBR connections can be characterized by Peak Cell Rate (PCR), Sustainable Cell Rate (SCR), and Maximum Burst Size (MBS). The delayed packets can be specified by Maximum Cell Transfer Delay (MaxCTD) parameter.

The Non Real Time -VBR (NRT-VBR) service category is used for the connections that transport traffic at variable bit rate and there is no inherent reliance on time synchronization between the traffic source and destination, but there is a need for an attempt to guarantee a specified bandwidth or latency. The VBR traffic grants some advantages as compared to the constant bit rate (CBR) traffic, as it offers higher quality and greater chance for statistical multiplexing [10-12].

3. VBR Traffic Parameters VBR traffic can be configured using four traffic parameters PCR, SCR, MaxCTD, and MBS [13-15]. The PCR and the SCR are the upper limit and the average of bits of data on the source traffic that can be sent, respectively. The MaxCTD is a parameter that determines the packet maximum delay in the network.

Whereas, the MBS defines the amount of time or the duration at which the source sends at the PCR.

VBR Traffic Forms VBR traffic can be formed in three different forms that are high traffic rate, sustainable traffic rate, and burst traffic rate. The sustainable traffic rate form is used when there are fewer activities, so the source transmits the data at the rate or near the rate of the SCR as shown in fig. 2.a. On the other hand, the high traffic rate form is used when there are high activities, so the source transmits the data at the rate or near the rate of the PCR as shown in fig. 2.b. In contrast, the burst traffic rate form is used when there are continual changes in the activities, so the source transmits the data at the rate between the SCR and PCR as illustrated in fig. 2.c.

4. THE PROPOSED ANALYTICAL MODEL

An analytical model to obtain the distribution of the number of bursts created by the assembly algorithm using both time- and threshold-based strategies was proposed in [16]. The model assumes that the packet arrival process is Poisson. The resulting burst process is characterized as a bulk burst transmission that is slotted in time.

Where, T is the length of the aggregation period, is the rate of the Poisson arrival process, n is the number of packets that arrive in T, the packet size is exponentially distributed with a mean of 1/b byte, avgNumPacks is equal to T, and s is equal to the square root of T.

This model was developed from two sub-models, which are the packets number and the number of bytes models, given respectively by:

The drawback of this model is the limited variability of the number of bytes in the assembly time T. This inability does not exist in the VBR traffic n the OBS network because of the number of bytes in T which can be clearly declared due to the characteristics of the VBR traffic.

Thus, the analytical model can give the required results by specifying the exact number of bytes in the aggregation time T. In the case of VBR traffic, the traffic parameters can easily determine the number of bytes in T. Thus, if the nodes are sending at a PCR rate, the proposed analytical model will be as follow:

5. RESULTS AND DISCUSSIONS

To validate the proposed analytical model, the number of bursts created in a period of time T produced by the Proposed analytical model and the results produced from simulation are compared in both cases SCR sending rates.

Similarly, figure 4 illustrates the comparison between the analytical results produced by the proposed model with the simulation results in the case where the nodes are sending VBR traffic with PCR rate. Once again, both results give identical values. The above figures (fig 3 and 4) show clearly that the results obtained from the proposed analytical model are in accordance with the results obtained from extensive simulations. This agreement in results validated the proposed analytical model.

6. CONCLUSION

In this paper, an analytical model of the number of bursts to be aggregated in a period of time for OBS VBR traffic is proposed. The model considers sending the VBR traffic is both PCR and SCR rates. To validate the proposed analytical model, the results produced by the model and the results produced by extensive simulations were compared. Both results are in accordance in the two cases (PCR and SCR).

REFERENCES

[1] I. Chlamtac, A. Ganz, and G. Karmi. "Lightpath communications: an approach to high-bandwidth optical WANs", IEEE Transactions on Communications, vol.40, pp.1171-1182, July 1992.

[2] D. Blumenthal. P. Prucnal, and J. Sauer. "Photonic packet switches - architectures and experimental implementations", IEEE 1667, November 1994. Proceedings, vol.82, 1650-

[3] C. Qiao and M. Yoo, "Optical Burst Switching (OBS) - a new paradigm for an optical Internet, " Journal High Speed Networks, vol. 8, pp. 69-84, 1999.

[4] J. Turner, "Terabit burst switching," Journal High Speed Networks, vol. 8, pp. 3-16, 1999.

[5] G. Hu, D. K., and C. Gauger, "Does burst assembly really reduce the self-similarity," in In Proc. of the Optical Fiber Communication Conference (OFC), 2003.

6] Y. Xiong, M. Vandenhoute, and H. Cankaya, "Control architecture in optical burst switched wdm networks", in IEEE- JSAC, vol. 18, no. 10, pp. 1838-1851,.2000,

[7] F. Xue and S. J. B. Yoo, "Self-similar traffic shaping at the edge router in optical packet-switched networks," In Proc. of IEEE International Conference on Communications, 2002.

[8] X. Yu, J. Li, X. Cao, Y. Chen, and C. Qiao, "Traffic statistics and performance evaluation in optical burst switched networks," Journal of lightwave technology, vol. 22, no. 12, pp. 2722-2738, 2004.

[9] ATM Forum, Traffic Management Specification, Version 4.0, April 1996, ftp://ftp.atmforum.com/pub/

[10] Gringeri, S., Shuaib, K., Egorov, R., Lewis, A., Khasnabish, B., and Basch, B. Traffic Shaping, Bandwidth Allocation, and Quality Assessment for MPEG Video Distribution over Broadband Networks. IEEE Network, 6 (Nov/Dec 1998), 94-107.

[11] Lakshman, T. V., Ortega, A., and Reibman, A. R. VBR : Tradeoffs and Potentials. Proceedings of the IEEE, 86, 5 (May 1998), 952-973.

[12] D. Lucantoni, M. Neuts, and A. Reibman, "Methods for performance evaluation of VBR video traffic models," IEEE/ACM Transactions on 176-180, April 1994. Networking, vol. 2, pp.

[13] ATM Forum, UNI 3.1 Signaling 3.1, September 1994, ftp://ftp.atmforum.com/pub/.

[14] ATM Forum, UNI Signaling 4.0, October 1995, ftp://ftp.atmforum.com/pub/.

[15] M. Hamdi and J. W. Roberts, "QoS guaranty for shaped bit rate video connections in broadband networks," in Proc. Int. Conf. Multimedia Networking (MmNet'95), Aizu-Wakamatsu, Japan, Oct. 1995.

[16] X. Mountrouidou "An Analytical Model of a Burst Networks" presented at HONET 2006, Charlotte, NC, Assembly Algorithm for Optical Burst Switched Sept. 6 - Sept. 8, 2006.

Faculty of IST, Multimedia University, Jalan Ayer Keroh Lama, 75450 Bukit Beruang, Melaka, Malaysia. Faculty of IT, Multimedia University, Jalan Multimedia, 63100 Cyberjaya, Malaysia mohammed.abdulatef05@mmu.edu.my
COPYRIGHT 2013 Asianet-Pakistan
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Al-Shargabi, M.A.; Abid, A.; Mellah, H.
Publication:Science International
Article Type:Report
Geographic Code:9PAKI
Date:Dec 31, 2013
Words:2668
Previous Article:INVESTIGATING THE PERFORMANCE OF SCHEDULING SCHEMES IN HSDPA SYSTEM.
Next Article:PROFILING AND ZONING OF GEOTECHNICAL SUB-SOIL DATA USING GEOGRAPHIC INFORMATION SYSTEM.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters