Скачать презентацию Development of an Automated Real-Time Performance Measurement and Скачать презентацию Development of an Automated Real-Time Performance Measurement and

dba311157cb257050c931a62cfeb7415.ppt

  • Количество слайдов: 33

Development of an Automated Real-Time Performance Measurement and Visualization Framework for Mobile Ad-hoc Networks Development of an Automated Real-Time Performance Measurement and Visualization Framework for Mobile Ad-hoc Networks BY: Shuprabha Shakya Dwijendra Kumar Das Dr. Yasuo Tsuchimoto Prof. Kanchanasut School of Engineering & Technologies, Asian Institute of Technology

shupa 643@gmail. com Presentation Outline Background § Background Problem § Current Problem Objectives Methodology shupa 643@gmail. com Presentation Outline Background § Background Problem § Current Problem Objectives Methodology System Benefits & Evaluation § Research Concept / Objectives § Design and Implementation § System Benefits and Evaluation § Conclusion and Future Work Conclusion An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com Background Problem Objectives Methodology System Benefits & Evaluation Conclusion § Mobile shupa 643@gmail. com Background Problem Objectives Methodology System Benefits & Evaluation Conclusion § Mobile Ad-hoc network is a promising approach for next generation wireless network. § Mobile wireless network are highly susceptible to various parameters like mobility, dynamicity of the environment, nodes distribution, variable available bandwidth etc. § Can be deployed in various scenario like collaborative computing, emergency network like disaster area etc. § Example – DUMBO [www. interlab. ait. ac. th/dumbo] [1] § Under emergency situation, network tends to be highly unpredictable and fragile. An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com Background What is the cause of deviation ? ? Background Problem shupa 643@gmail. com Background What is the cause of deviation ? ? Background Problem What is the current performance of my node? I wish to tune my system at real-time so as to sustain communication. . . Objectives Disaster site Methodology System Benefits & Evaluation Conclusion Under highly unpredictable circumstances, Real-time Performance information holds a great significance in order to maintain proper communication. An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com Background Problem Objectives Methodology System Benefits & Evaluation Conclusion Currently Used shupa 643@gmail. com Background Problem Objectives Methodology System Benefits & Evaluation Conclusion Currently Used Tools § Ethereal, Wireshark [8] § Tcpdump [9] § Dsniff [10] § m. Lab [3] § Offline !! An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com Current Approach Background Design Network & collect trace Problem Objectives During shupa 643@gmail. com Current Approach Background Design Network & collect trace Problem Objectives During Test Post Test Methodology System Benefits & Evaluation Maintain Huge Trace Record Offline Processing Data Analysis Phase Conclusion Final Results An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com What are the Problems? Background § Prolonged offline post processing (cannot shupa 643@gmail. com What are the Problems? Background § Prolonged offline post processing (cannot identify cause of deviation instantly) – not effective under unpredictable scenario e. g. Disaster Network Problem § Time Consuming Objectives § Complex for novice operator or non-programmer analyst Methodology System Benefits & Evaluation Conclusion § High memory storage requirement, unfavorable for long term experiment esp. under limited resource constraints. § Lack of a system capable of providing user with real-time performance showing dynamic affect of the changing environments Realized in DUMBO event An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com Research Concept / Objectives Background Problem Objectives Methodology System Benefits & shupa 643@gmail. com Research Concept / Objectives Background Problem Objectives Methodology System Benefits & Evaluation To develop a robust system for performing performance measurement and analysis at real time. § Study different components, their role and inter-relationship. § Design and develop a robust system for performing performance measurement and analysis at real time. § To integrate the system into routing protocol and provide realtime graphical visualization. § To validate the system with respect to the existing tools. Conclusion An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com System Requirements Background At Real- time, Problem § Distinguish between different shupa 643@gmail. com System Requirements Background At Real- time, Problem § Distinguish between different packet flows and extract information specific to the protocol. Objectives § Perform automated calculation based on the captured traffic. Methodology System Benefits & Evaluation Conclusion § Provide with the detail information for analysis along with summarized output. § Provide graphical performance monitoring tool for detecting the network performance. An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com System Requirements Background Problem Objectives § Keep track of network performance shupa 643@gmail. com System Requirements Background Problem Objectives § Keep track of network performance variation and trigger user of the sudden change in the network as it exceeds the threshold value. § Maintain the minimal calculated trace logs for the further analysis to overcome huge storage requirement. Methodology System Benefits & Evaluation § System should be easily customizable and easy to integrate with different MANET protocols. Conclusion An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com System (RTPMt) Design Background Problem Objectives Packets to Network Interface Ethernet shupa 643@gmail. com System (RTPMt) Design Background Problem Objectives Packets to Network Interface Ethernet Card Packet Capturing and Extraction Unit MANET Packet Format Packet Decoding Unit Methodology System Benefits & Evaluation Validation Controller Protocol Daemon Real Time Processing Unit Neighbor Table Hello Interval Real time Performance Analysis parameters Conclusion Real time trace and logs Trigger Detector An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com System Integration with Routing Protocol Background Problem Objectives Methodology Mobile Node shupa 643@gmail. com System Integration with Routing Protocol Background Problem Objectives Methodology Mobile Node Application Load Plug-in “ Real-Time Performance Measurement System ” Protocol Daemon Fetch functions and variables acess -RTPMt System Benefits & Evaluation Conclusion • Plug-in remains isolated – Flexible and extensible easily. • Cross Layer Design Advantage An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com Measurement Parameters Background § Packet Loss / Packet Loss Rate / shupa 643@gmail. com Measurement Parameters Background § Packet Loss / Packet Loss Rate / Packet Delivery Ratio Problem § Message Loss / Message Loss Rate Objectives Methodology § Control Overhead § Link Failure / Link Failure Frequency System Benefits & Evaluation Conclusion An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com Measurement Parameters – What? § Avg. Packet Loss : § Avg. shupa 643@gmail. com Measurement Parameters – What? § Avg. Packet Loss : § Avg. Packet Loss Rate : § Background Avg. PDR : Problem Objectives § Avg. Message Loss : § Avg. Message Loss Rate : § Avg. Control Overhead : § Avg. Link Failure Frequency : Methodology System Benefits & Evaluation Conclusion Skip Defined according to RFC 3626 [6] An Automated Real-Time Performance Measurement and Visualization Framework for MANET

Software and Platform Used Background § Platform: Problem Objectives § Tools and Library: Methodology Software and Platform Used Background § Platform: Problem Objectives § Tools and Library: Methodology Linux (Fedora Core or Ubuntu) Cygwin (for cross compilation in Win 32 platform) GNU C (gcc) Gtk +-2. 0 Libglade 2. 0 System Benefits & Evaluation § Protocol Used for study Conclusion Optimized Link State Routing Protocol (OLSR) [4] An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com System Overview Background Six Sections : § Input parameter section Problem shupa 643@gmail. com System Overview Background Six Sections : § Input parameter section Problem § Output parameter section Objectives § Sniffer section Methodology § Graph Plot section System Benefits & Evaluation § Analysis Section Conclusion § Help Section An Automated Real-Time Performance Measurement and Visualization Framework for MANET

Plug-in – Sniffer Section Background Problem Objectives Methodology System Benefits & Evaluation Conclusion An Plug-in – Sniffer Section Background Problem Objectives Methodology System Benefits & Evaluation Conclusion An Automated Real-Time Performance Measurement and Visualization Framework for MANET

Sniffer Section Background Problem Objectives Sniffer : Detail Packet Information Methodology System Benefits & Sniffer Section Background Problem Objectives Sniffer : Detail Packet Information Methodology System Benefits & Evaluation Conclusion An Automated Real-Time Performance Measurement and Visualization Framework for MANET

Plug-in – Sniffer Section Background Problem Objectives Methodology System Benefits & Evaluation Conclusion An Plug-in – Sniffer Section Background Problem Objectives Methodology System Benefits & Evaluation Conclusion An Automated Real-Time Performance Measurement and Visualization Framework for MANET

Output Parameter Section Background Problem Objectives Output: Summarized Traffic Flows and Statistics Methodology System Output Parameter Section Background Problem Objectives Output: Summarized Traffic Flows and Statistics Methodology System Benefits & Evaluation Conclusion An Automated Real-Time Performance Measurement and Visualization Framework for MANET

Plug-in – Sniffer Section Background Problem Objectives Methodology System Benefits & Evaluation Conclusion An Plug-in – Sniffer Section Background Problem Objectives Methodology System Benefits & Evaluation Conclusion An Automated Real-Time Performance Measurement and Visualization Framework for MANET

Graphical Visualization Section Background Problem Objectives Graph Plot: Graphical Visualization Methodology System Benefits & Graphical Visualization Section Background Problem Objectives Graph Plot: Graphical Visualization Methodology System Benefits & Evaluation Conclusion An Automated Real-Time Performance Measurement and Visualization Framework for MANET

Plug-in – Sniffer Section Background Problem Objectives Methodology System Benefits & Evaluation Conclusion An Plug-in – Sniffer Section Background Problem Objectives Methodology System Benefits & Evaluation Conclusion An Automated Real-Time Performance Measurement and Visualization Framework for MANET

Comparative Analysis Section Background Problem Objectives Comparative Evaluation : Diagnosis Methodology System Benefits & Comparative Analysis Section Background Problem Objectives Comparative Evaluation : Diagnosis Methodology System Benefits & Evaluation Conclusion An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com Sample Tests and Results Background Problem A 300 secs Objectives Methodology shupa 643@gmail. com Sample Tests and Results Background Problem A 300 secs Objectives Methodology Avg. Packet Loss Rate System Benefits & Evaluation Conclusion An Automated Real-Time Performance Measurement and Visualization Framework for MANET

Sample Tests and Results Avg. PDR = (received packet / transmitted packets) Background Problem Sample Tests and Results Avg. PDR = (received packet / transmitted packets) Background Problem System can reflect dynamic change in scenario Objectives Methodology (secs) Can adapt to establish robust network (secs) Avg. Link Failure Frequency System Benefits & Evaluation Conclusion (secs) An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com Sample Tests and Results Background System Trigger Problem Objectives Methodology System shupa 643@gmail. com Sample Tests and Results Background System Trigger Problem Objectives Methodology System is interactive System Benefits & Evaluation Conclusion An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com System Evaluation Background Problem § For validation purpose, RTPMt is compared shupa 643@gmail. com System Evaluation Background Problem § For validation purpose, RTPMt is compared with industry standard tool Ethereal. Objectives § Evaluation Parameters Methodology System Benefits & Evaluation Conclusion Qualitative Evaluation Quantitative Evaluation Time -Test Duration Distance Mobility - Number of mobile Nodes Node Density - Number of Nodes § On comparing, comparable results were obtained where the maximum deviation in performance is shown up to 0. 1 An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com Problems - System Benefit Background Problem § (Time Consuming) Instant information shupa 643@gmail. com Problems - System Benefit Background Problem § (Time Consuming) Instant information at real-time, more valuable under highly unpredictable wireless environment § (Complex) Simplicity and Comprehensibility Summarization at different level of detailing according to the necessity Graphical visualization provides effective monitoring and comprehensible perception. Interactive and diagnostic feature (Run time trigger, customization of plot, comparative traffic analysis) Objectives Methodology System Benefits & Evaluation Conclusion § (Long Duration Test Limitation) Favors long time test in high node density; Reduction in the storage memory size requirement [11, Report] Tradeoff : § Cannot provide detailed information An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com Conclusion Background Problem Objectives Methodology System Benefits & Evaluation Conclusion § shupa 643@gmail. com Conclusion Background Problem Objectives Methodology System Benefits & Evaluation Conclusion § Designed and developed a robust real-time performance measurement system very useful for emergency network. § Real-time visualization and network performance summarization provides user with clear perception of network performance instantly. § Addresses problems and propose solution in implementing realtime approach [11]. § Robust ; Supports Long duration test with high node density due to reduction in storage requirement. § Design is flexible. An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com Future Work Background Problem Objectives § Current approach, limited to OLSR shupa 643@gmail. com Future Work Background Problem Objectives § Current approach, limited to OLSR can be extended for multiple protocols. § Further work is under progress for performance evaluation of the tools. Based on the result, our focus is to tune our system with respect to the standard tool so as to improve its efficiency. Methodology System Benefits & Evaluation Conclusion § System can be further standardized for effective measurement and evaluation in the future. § Features like real-time change in parameters can further be utilized to improve the protocol in adaptive way through real-time analysis. An Automated Real-Time Performance Measurement and Visualization Framework for MANET

shupa 643@gmail. com References Background Problem Objectives Methodology System Benefits & Evaluation Conclusion 1. shupa 643@gmail. com References Background Problem Objectives Methodology System Benefits & Evaluation Conclusion 1. Inter. Lab, Digital Ubiquitous Mobile Ad-hoc Network, DUMBO 2006. Asian Institute of Technology, Thailand, 2006. http: //www. interlab. ait. ac. th/dumbo 2. Karygiannis, A. and Antonakakis, E. m. Lab: A Mobile Ad Hoc Network Test Bed. 1 st Workshop on Security, Privacy and Trust in Pervasive and Ubiquitous Computing in conjunction with the IEEE International Conference in Pervasive Services 2005, July 14, 2005. 3. Marques, P. Castro, H. Ricardo, M. Monitoring emerging IPv 6 wireless access networks. Wireless Communication, IEEE, 2005. 4. Clausen T. , Jacquet P. , Laouiti A. , Minet P. , Muhlethaler P. , Qayyum A. & Viennot L. , Optimized Link State Routing Protocol(OLSR), IETF RFC 3626, 2005. 5. Perkins C. E. , Belding-Royer E. M. , and Das S. R. Ad-hoc On-Demand Distance Vector (AODV) Routing , IETF RFC 3561, July 2003. 6. Subbarao Madhavi W. , Ad Hoc Networking Critical Features and Performance Metric. Wireless Communications Technology Group, NIST, October 1999. 7. Schmitz P. and Weaver G. MIPv 6: New Capabilities for Seamless Roaming Among Wired, Wireless, and Cellular Networks, Developer UPDATEMagazine, Intel, September 2002. 8. Ethereal Network Protocol Analyzer. http: //www. ethereal. com/download. html 9. Tcp. Dump Network Monitoring Tool. http: //www. tcpdump. org/ 10. DSniff http: //www. monkey. org/~dugsong/dsniff/ 11. Shakya S. , Kanchanasut K. , Tsuchimoto Y. and Saengudomlert P. Development of automated real time performance measurement and visualization framework for real mobile ad-hoc network. Thesis Report, Asian Institute of Technologies, AIT, May 2007. An Automated Real-Time Performance Measurement and Visualization Framework for MANET

Background Problem Objectives Methodology System Benefits & Evaluation Conclusion An Automated Real-Time Performance Measurement Background Problem Objectives Methodology System Benefits & Evaluation Conclusion An Automated Real-Time Performance Measurement and Visualization Framework for MANET