` Research 2: Information Diversity through Information Flow Subgoal: Systematically and precisely measure program diversity by measuring the information.

Slides:



Advertisements
Similar presentations
A Hierarchical Multiple Target Tracking Algorithm for Sensor Networks Songhwai Oh and Shankar Sastry EECS, Berkeley Nest Retreat, Jan
Advertisements

UNIVERSITY OF JYVÄSKYLÄ Building NeuroSearch – Intelligent Evolutionary Search Algorithm For Peer-to-Peer Environment Master’s Thesis by Joni Töyrylä
Introduction to Cyber Physical Systems Yuping Dong Sep. 21, 2009.
Chapter 12: Expert Systems Design Examples
1 Sensor Networks and Networked Societies of Artifacts Jose Rolim University of Geneva.
E NERGY -A WARE G ENETIC A LGORITHM FOR M ANAGING W IRELESS S ENSOR N ETWORKS Abhishek Karpate.
SOFTWARE SECURITY JORINA VAN MALSEN 1 FLAX: Systematic Discovery of Client-Side Validation Vulnerabilities in Rich Web Applications.
1 Simulation Modeling and Analysis Session 13 Simulation Optimization.
Modern Information Retrieval
Configuration Management IACT 918 July 2004 Gene Awyzio SITACS University of Wollongong.
Genetic Algorithms Nehaya Tayseer 1.Introduction What is a Genetic algorithm? A search technique used in computer science to find approximate solutions.
Configuration Management IACT 418/918 Autumn 2005 Gene Awyzio SITACS University of Wollongong.
UNIVERSITY OF JYVÄSKYLÄ Resource Discovery in Unstructured P2P Networks Distributed Systems Research Seminar on Mikko Vapa, research student.
Attack-Resistant Networks Allen G. Taylor Communication networks have four primary objectives: Minimize.
Software Process and Product Metrics
1 Software Testing Techniques CIS 375 Bruce R. Maxim UM-Dearborn.
Subgoal: conduct an in-depth study of critical representation, operator and other choices used for evolutionary program repair at the source code level.
UML and Object Oriented Concepts
Image Registration of Very Large Images via Genetic Programming Sarit Chicotay Omid E. David Nathan S. Netanyahu CVPR ‘14 Workshop on Registration of Very.
For 3-G Systems Tara Larzelere EE 497A Semester Project.
Alternative Parallel Processing Approaches Jonathan Sagabaen.
1 Chap 10 Malicious Software. 2 Viruses and ”Malicious Programs ” Computer “Viruses” and related programs have the ability to replicate themselves on.
COGNITIVE RADIO FOR NEXT-GENERATION WIRELESS NETWORKS: AN APPROACH TO OPPORTUNISTIC CHANNEL SELECTION IN IEEE BASED WIRELESS MESH Dusit Niyato,
Development in hardware – Why? Option: array of custom processing nodes Step 1: analyze the application and extract the component tasks Step 2: design.
Revolutionizing the Field of Grey-box Attack Surface Testing with Evolutionary Fuzzing Department of Computer Science & Engineering College of Engineering.
Michael Ernst, page 1 Collaborative Learning for Security and Repair in Application Communities Performers: MIT and Determina Michael Ernst MIT Computer.
VeriFlow: Verifying Network-Wide Invariants in Real Time
1 Paper Review for ENGG6140 Memetic Algorithms By: Jin Zeng Shaun Wang School of Engineering University of Guelph Mar. 18, 2002.
Lecture #12 Distributed Algorithms (I) CS492 Special Topics in Computer Science: Distributed Algorithms and Systems.
Computer Science Open Research Questions Adversary models –Define/Formalize adversary models Need to incorporate characteristics of new technologies and.
Helix Automatic Software Repair with Evolutionary Computation Stephanie Forrest Westley Weimer.
Clone-Cloud. Motivation With the increasing use of mobile devices, mobile applications with richer functionalities are becoming ubiquitous But mobile.
Study on Genetic Network Programming (GNP) with Learning and Evolution Hirasawa laboratory, Artificial Intelligence section Information architecture field.
CSC-682 Cryptography & Computer Security Sound and Precise Analysis of Web Applications for Injection Vulnerabilities Pompi Rotaru Based on an article.
Design of a real time strategy game with a genetic AI By Bharat Ponnaluri.
1. Process Gather Input – Today Form Coherent Consensus – Next two months.
KNOWLEDGE BASED TECHNIQUES INTRODUCTION many geographical problems are ill-structured an ill-structured problem "lacks a solution algorithm.
Fuzzy Genetic Algorithm
Evolving Virtual Creatures & Evolving 3D Morphology and Behavior by Competition Papers by Karl Sims Presented by Sarah Waziruddin.
Simulation is the process of studying the behavior of a real system by using a model that replicates the behavior of the system under different scenarios.
Communication Paradigm for Sensor Networks Sensor Networks Sensor Networks Directed Diffusion Directed Diffusion SPIN SPIN Ishan Banerjee
REPRESENTATIONS AND OPERATORS FOR IMPROVING EVOLUTIONARY SOFTWARE REPAIR Claire Le Goues Westley Weimer Stephanie Forrest
Xusheng Xiao North Carolina State University CSC 720 Project Presentation 1.
1. Genetic Algorithms: An Overview  Objectives - Studying basic principle of GA - Understanding applications in prisoner’s dilemma & sorting network.
SPYCE/May’04 coverage: A Cooperative Immunization System for an Untrusting Internet Kostas Anagnostakis University of Pennsylvania Joint work with: Michael.
HNDIT23082 Lecture 06:Software Maintenance. Reasons for changes Errors in the existing system Changes in requirements Technological advances Legislation.
CITS7212: Computational Intelligence An Overview of Core CI Technologies Lyndon While.
Coevolutionary Automated Software Correction Josh Wilkerson PhD Candidate in Computer Science Missouri S&T.
CSCI1600: Embedded and Real Time Software Lecture 33: Worst Case Execution Time Steven Reiss, Fall 2015.
Neural Networks Presented by M. Abbasi Course lecturer: Dr.Tohidkhah.
Automated Worm Fingerprinting Authors: Sumeet Singh, Cristian Estan, George Varghese and Stefan Savage Publish: OSDI'04. Presenter: YanYan Wang.
Detecting, Managing, and Diagnosing Failures with FUSE John Dunagan, Juhan Lee (MSN), Alec Wolman WIP.
Design of a real time strategy game with a genetic AI By Bharat Ponnaluri.
Dynamic Load Balancing Tree and Structured Computations.
Presentation By SANJOG BHATTA Student ID : July 1’ 2009.
Multi-cellular paradigm The molecular level can support self- replication (and self- repair). But we also need cells that can be designed to fit the specific.
` Question: How do immune systems achieve such remarkable scalability? Approach: Simulate lymphoid compartments, fixed circulatory networks, cytokine communication.
A Presentation on Adaptive Neuro-Fuzzy Inference System using Particle Swarm Optimization and it’s Application By Sumanta Kundu (En.R.No.
Genetic Algorithm. Outline Motivation Genetic algorithms An illustrative example Hypothesis space search.
Vigilante: End-to-End Containment of Internet Worms Manuel Costa, Jon Crowcroft, Miguel Castro, Antony Rowstron, Lidong Zhou, Lintao Zhang and Paul Barham.
Design Patterns: MORE Examples
Software Testing.
Scaling for the Future Katherine Yelick U.C. Berkeley, EECS
CSCI1600: Embedded and Real Time Software
Digital Processing Platform
Software Testing (Lecture 11-a)
Lecture 4. Niching and Speciation (1)
CSCI1600: Embedded and Real Time Software
Coevolutionary Automated Software Correction
Presentation transcript:

` Research 2: Information Diversity through Information Flow Subgoal: Systematically and precisely measure program diversity by measuring the information flow generated by unique inputs. Insight: Attacks and defects have unique information flow signatures. Conversely, bug fixes exhibit information flows that differ in a significant manner from the original program Status: Status: Dynamic Information Flow Tracking infrastructure is complete and is being unit tested, x86 instruction set is 15% instrumented. Approach: Extend our existing platforms by further examining biological factors Systems contain more errors and are more prone to attack than ever. The balance of power favors the attacker:  Software replicates are all vulnerable to the same attack.  System complexity precludes rapid repair. We must rethink the current cybersecurity paradigm. Subgoal: Develop models and simulations to understand Scalable RADAR principles and adapt them to computation, specifically to improve our existing techniques. Scalable RADAR for Co-evolutionary Adaptive Environments Wes Weimer Stephanie Forrest (PI)Jed CrandallMelanie Moses Biological systems search complex spaces without a “leader.” Biological Principles Insight Cells respond to environmental signals automatically. Research 1: Evolutionary Program Repair Mutational Robustness and Proactive Diversity process and attempt to fix 105 indicative bugs found in existing programs. Subgoal: Apply evolutionary repair to known bugs in real-world programs totaling over 5 million lines of code and 10,000 test cases. Problem Animal immune systems can defeat multiple, adaptable adversaries. Adapt Scalable RADAR to a new, clean-slate paradigm for software development/maintenance. Demonstrate large, complex software systems that:  automatically detect attacks  repair themselves  evolve a diversity of solutions. Goal Evaluating Diversity Distributed Repair Study of the Immune System Immune systems are composed of millions of cells. Redundancy, diversity, “wisdom of the crowd.” Genes, cells, systems adapt over multiple time scales. Software is a complex, evolving system. Research 3: Simulation and Modeling ProgramLOCRepaired / Total fbc 97k1 / 3 gmp 145k1 / 2 gzip 491k1 / 5 libtiff 77k17 / 24 lighttpd 62k5 / 9 php 1,046k28 / 44 python 407k1 / 11 wireshark 2,814k1 / 7 total5,139k55 / 105 Systematic Study of Cost and Generality Subgoal 1: Examine whether there is a computational analog for biological mutational robustness and thus quantify the ability of random changes to produce variants that retain specified program behavior. Subgoal 2: Use mutational robustness to proactively fix unknown bugs while retaining functionality. Question 1: How much do inflammatory search signals speed up immune repair? Answer: Biologically, as the size of the search space increases, the effect of signals improves performance by orders of magnitude. Question 2: How much does the structure of the lymphatic system speed up repair? Answer: There is a trade- off between many small nodes and few large nodes – rate of distribution of repairs vs. speed of recruitment of new repairs. We replicate optimal immune network structure to distribute repairs. Question: What effect does diversity have on vulnerability? Answer: Diversity decreases with increased connectivity and communication. Diversity decreases network vulnerability, even when it increases individual vulnerability. Example: Despite larger individual vulnerabilities (in red),The group AB above is less vulnerable than CDE Question: Can our existing approach be adapted to repair specialized embedded devices? Approach: While individual devices may lack the computing speed to efficiently find repairs, they can cooperate to explore the search space and find repairs more quickly Future: Simulate spread of both attacks and repairs simultaneously Hello0x2 0 There\0 Input[Loop] != Input[Loop] >= ‘a’ Input[Loop] <= ‘z’ Input[Loop] >= ‘A’ Input[Loop] <= ‘Z’ Approach: Construct matrices (pictured above) relating input to branch decisions. Judge the diversity of programs by comparing their structure in a way that is robust to small, simple changes for(Loop = 0; Input[Loop] != ‘\0’; Loop++){ … if ((Input[Loop] >= ‘a’) && (Input[Loop] <= ‘z’)) { … else if((Input[Loop] >= ‘a’) && (Input[Loop] <= ‘z’)) { … Approach: Enhance several fundamental steps throughout the Result: Improvements yielded 68% more patches. Based on Amazon EC2 cloud service rates, 55 bugs were fixed at an average cost of $7.32 per bug. Results: We select a population of variants based on computational analogs of biological diversity that fixes an average of 40% of unknown bugs. Mutational robustness: Independent of programming language, domain, and test suit coverage, the fraction of program variants with identical behavior on all available test cases is 36.75% in 22 programs.