NETWORK DELAY: HOW RELIABLE NETWORK ANALYZER SOFTWARE DEVELOPMENT

Transcription

1 NETWORK DELAY: HOW RELIABLE NETWORK ANALYZER SOFTWARE DEVELOPMENT 1 MOHD NAZRI ISMAIL, 2 ABDULLAH MOHD ZIN 1 Faculty of MIIT, University of Kuala Lumpur (UniKL), Malaysia 2 Faculty of FTSM, University of Kebangsaan Malaysia (UKM), Malaysia [email protected], [email protected] ABSTRACT This paper presents a complete network analyzer development for network delay in campus environment. The purpose of this study is to define the accuracy of network analyzer development with independent data, real network and OPNET simulation tool. The network delay will measure based on transmission delay and propagation delay. This network analyzer software will test on delay generated by the several services. The reliability of this network analyzer will test with , text messaging and instant messaging over web service. The results show that network analyzer software has accuracy and same trend delay with independent data, real network and OPNET simulation tool. Finally, this software is able to measure the network delay during preparation, proposal and planning phases. Keywords: Accuracy, Traffic, Analyzer, OPNET 1. INTRODUCTION This study focuses on the accuracy of network delay using heterogeneous services. This study does not intend to perform a comprehensive test the functionality of all simulator and analyzer features. OPNET has originally been developed for network simulation and it is fully usable as a robust and reliability simulation tool with higher investment. This network analyzer development process has discussed detail in [1], [2], [3]. Table 1.1 shows service performance requirement for several services. Service performance requirement consists of delay, capacity (bandwidth) and reliability [4]. Reliability is a measure of the network/system ability to provide deterministic and accurate delivery of information [5], [6]. Delay is a time difference in transmitting a single unit of information (bit, byte, cell, frame, and packet) from source to destination. Table 1.1 Service Performance Requirements 2. DEFINITION OF NETWORK ANALYZER A network analyzer also called a "packet analyzer," "traffic analyzer" and "protocol analyzer," [7]. analyzers functionality such as [8]: i) provide detailed statistics for current and recent activity on the network; ii) detect unusual levels of network traffic; iii) detect unusual packet characteristics; iv) identify packet sources or destinations; v) configure alar for defined threats; vi) and vii) monitor bandwidth utilization. Management: management consists of a variety of tasks, for example, monitoring, configuration, troubleshooting and planning that are performed by users and network administrators [9]. element is a component of the network that can be managed. This includes hosts, routers, switches, hubs and server that can be measured. Examples of end-toend characteristics for network elements and network traffic are capacity (bandwidth), availability, delay, jitter, throughput, network utilization and error rates [10], [11]. 3. METHODOLOGY Figure 3.1 shows network life cycle approach for technologies and services implementation in the future [12]. life cycle approach consists 9

2 of six phases such as prepare, plan, design, implement, operate and optimize. This network analyzer development concentrates more on preparation, planning and proposal areas. Figure 3.1: Comparison of Analyzer, OPNET and Real analyzer development is based on mathematical model. We use queuing theory M/M/1 to build this software [13], [14]. This software was developed to measure and plan network activities such as predict usage of network delay. (refer to Figure 3.2). The reliability test will only concentrate on network delay. Figure 3.3: Analyzer Accuracy Testing Process Figure 3.4: Input Parameters for Analyzer, OPNET and Real 4. ANALYSIS AND RESULTS Figure 3.2: Analyzer Modules Development Figure 3.3 shows network delay reliability test. The independent data output is generated based on number of distance input, number of nodes input, size of bandwidth input and size of services input. These inputs will use in OPNET, real network and network analyzer software (refer to Figure 3.4). We conduct three experiments to confirm the reliability of network delay via independent data, OPNET application and real network. In our experiments, we will test , instant messaging and text messaging over web service. The network delay will measure based on transmission delay and propagation delay. Other delay will not consider in network analyzer development. Figure 4.1 show the network delay measurement performance via network analyzer software design, OPNET application design and 10

3 real network design. Table 4.1 shows input for independent data for single user that will be used in network analyzer software, OPNET application and real network [15]. The following diagram output shows network delay over different packet size for single user (refer to Figure 4.2, Figure 4.3 and Figure 4.4). Figure 4.1: Design for Reliability Test Table 4.1: Inputs for Independent Data to Evaluate Delay Figure 4.2: Delay for 100 Bytes Types of Services (Low) Instant Messaging (Low) Text Messaging (Low) Size of Services Estimation 1000 Bytes 500 Bytes 100 Bytes 4.1 Independent Data Delay In our network delay experiment, independent data is used to measure network delay (transmission and propagation delay) reliability between network analyzer development, OPNET application and real network. Project conducted by David Grangier, Institut Eurecom, France, is selected to use as independent data [16] (refer to Figure 4.1 and Table 4.2). Figure 4.3: Delay for 500 Bytes Figure 4.4: Delay for 1000 Bytes Figure 4.1: Sample of Independent Data Input Table 4.2: Independent Data Input Input to Measure Delay Length Rate (KM) (Mbps) and Delay (Propagation and Transmission Delay) Packet size (Bytes) 4.2 OPNET Application Delay Figure 4.3 shows OPNET design via different of packet size over local area network. Only one user will test on the reliability of network analyzer development with OPNET application and independent data. Size of , instant messaging and text messaging over web service will configure in OPNET application (refer to Figure 4.4, Figure 4.5 and Figure 4.6) then it will simulate the scenario and measure the network delay performance. 11

4 The results show that OPNET application generates as follow network delay: i) 100 bytes sec/ to sec/0.1; ii) 500 bytes sec/0.05 to sec/0.15; and iii) 1000 bytes sec/0.075 to sec/0.17 (refer to Figure 4.7, Figure 4.8 and Figure 4.9). Figure 4.3: OPNET Application: Delay Design Figure 4.4: Size of Text Messaging Configuration over Web Service Figure 4.7: Delay for Text Messaging Figure 4.5: Size of Instant Messaging Configuration over Web Service Figure 4.8: Delay for Instant Messaging (500 bytes) Figure 4.6: Size of Configuration over Web Service Figure 4.9: Delay for (1000 bytes) 12

5 4.3 Analyzer Development Delay Then, network analyzer development will configure using the same input from independent data and OPNET application. First experiment will configure 5 km and second experiments will configure 10 km. The network analyzer development shows network delay results for 5 km (refer to Figure 4.10) and 10 km (refer to Figure 4.11). Real network environment has configured to enable network delay over LAN. The , text messaging and instant messaging traffic will pump using fluke network analyzer to the server. management tool such as ColaSoft Capsa and Visualware application are used to capture network delay. Figure 4.12 shows real network traffic for 100 bytes use by text messaging service that has received by the client. Table 4.3, Table 4.4 and Table 4.5 show network delay results generate by Visualware as follow: i) 100 bytes 0; ii) 500 bytes 0 to 1 ; and iii) 1000 bytes 0 to 2. Figure 4.10: Delay for 5 KM Figure 4.14: Real Delay for 100 Bytes (Text Messaging) 4.4 Real Delay Figure 4.11: Delay for 10 KM Table 4.3: Delay Text Messaging (100 Bytes) 13

6 Hops Loss IP Name Avg Min Max lect-mohdnazri.unikl.edu.my [Local ] MIIT-ADIDAHLAJI [Local ] Table 4.4: Delay Instant Messaging (500 Bytes) Hops Loss IP Name Avg Min Max lect-mohdnazri.unikl.edu.my [Local ] [Local ] [Local ] Table 4.5: Delay (1000 Bytes) Hops Loss IP Name Avg Min Max lect-mohdnazri.unikl.edu.my [Local ] [Local ] [Local ] We conclude all our findings in Table 4.6 and Table 4.7. The results generate from network analyzer is closely resemble with OPNET application, real network and independent data. Again, it is confirm and proof that network analyzer development is able to predict and plan network delay usage for heterogeneous services. Table 4.6: Delay - Reliability of Analyzer Development between Opnet, Independent Data and Real Table 4.7: Delay - Reliability of Analyzer Development between Opnet, Independent Data and Real 5. CONCLUSION Today's networking environment has become very complex. s have been growing in size rapidly and support complex applications. Even, our network analyzer development can determine and solve proble for homogenous and heterogeneous services in LAN and WAN. The results show that network analyzer development is able to measure and analyze 14

7 approximately same as OPNET simulation tool, independent data and real network for network delay. This network analyzer development can use to measure and analyze network delay behavior for preparation and planning purposes. In addition, it is easy to use and provide a userfriendly graphical and text interface. REFRENCES: [1] M. Fleury, G. Flores Lucio, and M. J. Reed. Clarification of the OPNET NS-2 Comparison Paper with. regards to OPNET Modeler. ET-NS2_Comparison.pdf, accessed 9/9/2008. [2] Brown K, Christianson L, OPNET Lab Manual to Accompany Data and Computer Communications 7th ed. And Computer ing with Internet Protocols and Technology 4th ed. by William Stallings, Pearson Prentice Hall, Upper Saddle River, NJ, [3] Xinjie Chang Simulations With OPNET. Proceedings of the 1999 Winter Simulation Conference P. A. Farrington, H. B. Nembhard, D. T. Sturrock, and G. W. Evans, eds. [4] James D. McCabe. Practical Computer Analysis and Design, pp , 1998, Morgan Kaufmann Publishers. [5] Michael R. Lyu. Software Reliability Engineering: A Roadmap, International Conference on Software Engineering, pp , [6] Abdullah Konak. Combining network reductions and simulation to estimate network reliability, Proceedings of the 39th conference on Winter simulation, pp , [Oktober 11, 2009] [7] Victor A. Clincy & Nael Abu-Halaweh A Taxonomy of free Sniffers for teaching and research, Source Journal of Computing Sciences in Colleges Vol. 21(1), pp [8] J. P. Talledo Design and Implementation of an Ethernet Frame Analyzer for High Speed s, Proceedings of the 15th International Conference on Electronics, Communications and Computers, Publisher IEEE Computer Society pp [9] Jairo A. Gutiérrez. A connectionless approach to integrated network management, International Journal of Management Vol. 8(4), pp , 1998, John Wiley & Sons, Inc. New York, NY, USA [10] Faouzi Kamoun. Toward best maintenance practices in communications network management, International Journal of Management,Vol. 15(5), pp , 2005, John Wiley & Sons, Inc. New York, NY, USA [11] A.S. Sethi. Bibliography on network management, ACM SIGCOMM Computer Communication Review, Vol. 19(3), pp , [12] Cisco s Cisco s Life Cycle Services Approach, Cisco ers Conference 07, Januari 2007, Sunway Pyramid, Malaysia. [13] Mohd Nazri Ismail, Abdullah Mohd Zin, "Comparing the Accuracy of End-to-End Performance Measurement Testbed and Simulation Model for Data Transfers in Heterogeneous Environment," a, pp , IEEE, Second Asia International Conference on Modelling & Simulation (AMS), [14] Mohd Nazri Ismail and Abdullah Mohd Zin. Development of Simulation Model in Heterogeneous Environment: Comparing the Accuracy of Simulation Model for Data Transfers Measurement over Wide Area. Asian for Scientific Information. Information Technology Journal, pp. 2448, ISSN: (Print). Pakistan. [15] Hans Lohninger. Wireless ing in the Developing World, January 2007, [Online access], wndw_05_07.html [16] David Grangier. Transmission versus Propagation Delay, Eurecom Institute, [Online access], ~liu/kebin/applet1/delay.htm 15