Presentation is loading. Please wait.

Presentation is loading. Please wait.

Hamsa: Fast Signature Generation for Zero-day Polymorphic Worms with Provable Attack Resilience Zhichun Li, Manan Sanghi, Yan Chen, Ming-Yang Kao and Brian.

Similar presentations


Presentation on theme: "Hamsa: Fast Signature Generation for Zero-day Polymorphic Worms with Provable Attack Resilience Zhichun Li, Manan Sanghi, Yan Chen, Ming-Yang Kao and Brian."— Presentation transcript:

1 Hamsa: Fast Signature Generation for Zero-day Polymorphic Worms with Provable Attack Resilience Zhichun Li, Manan Sanghi, Yan Chen, Ming-Yang Kao and Brian Chavez Lab for Internet & Security Technology (LIST) Northwestern University

2 2 The Spread of Sapphire/Slammer Worms

3 3 Desired Requirements for Polymorphic Worm Signature Generation Network-based signature generation –Worms spread in exponential speed, to detect them in their early stage is very crucial… However »At their early stage there are limited worm samples. –The high speed network router may see more worm samples… But »Need to keep up with the network speed ! »Only can use network level information

4 4 Desired Requirements for Polymorphic Worm Signature Generation No existing work satisfies these requirements ! Noise tolerant –Most network flow classifiers suffer false positives. –Even host based approaches can be injected with noise. Attack resilience –Attackers always try to evade the detection systems Efficient signature matching for high-speed links

5 5 Outline Motivation Hamsa Design Model-based Signature Generation Evaluation Related Work Conclusion

6 6 Choice of Signatures Two classes of signatures –Content based »Token: a substring with reasonable coverage to the suspicious traffic »Signatures: conjunction of tokens –Behavior based Our choice: content based –Fast signature matching. ASIC based approach can archive 6 ~ 8Gb/s –Generic, independent of any protocol or server

7 7 Unique Invariants of Worms Protocol Frame –The code path to the vulnerability part, usually infrequently used –Code-Red II: ‘.ida?’ or ‘.idq?’ Control Data: leading to control flow hijacking –Hard coded value to overwrite a jump target or a function call Worm Executable Payload –CLET polymorphic engine: ‘0\x8b’, ‘\xff\xff\xff’ and ‘t\x07\xeb’ Possible to have worms with no such invariants, but very hard Invariants

8 8 Hamsa Architecture

9 9 Hamsa Design Key idea: model the uniqueness of worm invariants –Greedy algorithm for finding token conjunction signatures Highly accurate while much faster –Both analytically and experimentally –Compared with the latest work, polygraph –Suffix array based token extraction Provable attack resilience guarantee Noise tolerant

10 10 Hamsa Signature Generator Core part: Model-based Greedy Signature Generation Iterative approach for multiple worms

11 11 Outline Motivation Hamsa Design Model-based Signature Generation Evaluation Related Work Conclusion

12 12 Problem Formulation Signature Generator Signature false positive bound  Maximize the coverage in the suspicious pool False positive in the normal pool is bounded by  Suspicious pool Normal pool With noise NP-Hard!

13 13 Model Uniqueness of Invariants FP 21% 9% 17% 5% t1t1 Joint FP with t 1 2% 0.5% 1% t2t2 The total number of tokens bounded by k* U(1)=upper bound of FP( t 1 ) U(2)=upper bound of FP( t 1,t 2 )

14 14 Signature Generation Algorithm (82%, 50%) (COV, FP) (70%, 11%) (67%, 30%) (62%, 15%) (50%, 25%) (41%, 55%) (36%, 41%) (12%, 9%) u(1)=15% Suspicious pool tokens token extraction Order by coverage t1t1

15 15 (82%, 50%) (COV, FP) (70%, 11%) (67%, 30%) (62%, 15%) (50%, 25%) (41%, 55%) (36%, 41%) (12%, 9%) t1t1 Order by joint coverage with t 1 (69%, 9.8%) (COV, FP) (68%, 8.5%) (67%, 1%) (40%, 2.5%) (35%, 12%) (31%, 9%) (10%, 0.5%) u(2)=7.5% t2t2 Signature Signature Generation Algorithm

16 16 Algorithm Analysis Runtime analysis O(T*(|M|+|N|)) Provable Attack Resilience Guarantee –Analytically bound the worst attackers can do! –Example: K*=5, u(1)=0.2, u(2)=0.08, u(3)=0.04, u(4)=0.02, u(5)=0.01 and  =0.01 –The better the flow classifier, the lower are the false negatives Noise ratioFP upper boundFN upper bound 5%1%1.84% 10%1%3.89% 20%1%8.75%

17 17 Attack Resilience Assumptions Two Common assumptions for any sig generation sys Two Unique assumptions for token-based schemes Attacks to the flow classifier –Our approach does not depend on perfect flow classifiers –With 99% noise, no approach can work! –High noise injection makes the worm propagate less efficiently. Enhance flow classifiers

18 18 Improvements to the Basic Approach Generalizing Signature Generation –use scoring function to evaluate the goodness of signature Iteratively use single worm detector to detect multiple worms –At the first iteration, the algorithm find the signature for the most popular worms in the suspicious pool. –All other worms and normal traffic treat as noise.

19 19 Outline Motivation Hamsa Design Model-based Signature Generation Evaluation Related Work Conclusion

20 20 Experiment Methodology Experiential setup: –Suspicious pool: »Three pseudo polymorphic worms based on real exploits (Code-Red II, Apache-Knacker and ATPhttpd), »Two polymorphic engines from Internet (CLET and TAPiON). –Normal pool: 2 hour departmental http trace (326MB) Signature evaluation: –False negative: 5000 generated worm samples per worm –False positive: »4-day departmental http trace (12.6 GB) »3.7GB web crawling including.mp3,.rm,.ppt,.pdf,.swf etc. »/usr/bin of Linux Fedora Core 4

21 21 Results on Signature Quality Single worm with noise –Suspicious pool size: 100 and 200 samples –Noise ratio: 0%, 10%, 30%, 50%, 70% –Noise samples randomly picked from the normal pool –Always get above signatures and accuracy. Multiple worms with noises give similar results Worms Training FN Training FP Evaluation FN Evaluation FP Binary evaluation FP Signature Code-Red II 00000 {'.ida?': 1, '%u780': 1, ' HTTP/1.0\r\n': 1, 'GET /': 1, '%u': 2} CLET00.109%00.06236%0.268% {'0\x8b': 1, '\xff\xff\xff': 1,'t\x07\xeb': 1}

22 22 Speed Results Implementation with C++/Python –500 samples with 20% noise, 100MB normal traffic pool, 15 seconds on an XEON 2.8Ghz, 112MB memory consumption Speed comparison with Polygraph –Asymptotic runtime: O(T) vs. O(|M| 2 ), when |M| increase, T won’t increase as fast as |M|! –Experimental: 64 to 361 times faster (polygraph vs. ours, both in python)

23 23 Outline Motivation Hamsa Design Model-based Signature Generation Evaluation Related Work Conclusion

24 24 Related works HamsaPolygraphCFGPADSNemeanCOVERSMalware Detection Network or host based Network Host Content or behavior based Content based Behavior based Content based Behavior based Noise tolerance YesYes (slow) YesNo Yes Multi worms in one protocol YesYes (slow) YesNoYes On-line sig matching Fast SlowFast Slow GeneralityGeneral purpose Protocol specific Server specific General purpose Provable atk resilience YesNo Information exploited   

25 25 Conclusion Network based signature generation and matching are important and challenging Hamsa: automated signature generation –Fast –Noise tolerant –Provable attack resilience –Capable of detecting multiple worms in a single application protocol Proposed a model to describe the worm invariants

26 Questions ?

27 27 Experiment: Sample requirement Coincidental-pattern attack [Polygraph] Results –For the three pseudo worms, 10 samples can get good results. –CLET and TAPiON at least need 50 samples Conclusion –For better signatures, to be conservative, at least need 100+ samples Require scalable and fast signature generation!

28 28 Experiment: U-bound evaluation To be conservative we chose k * =15. –Even we assume every token has 70% false positive, their conjunction still only have 0.5% false positive. In practice, very few tokens exceed 70% false positive. Define u(1) and u r, generate –We tested: u(1) = [0.02, 0.04, 0.06, 0.08, 0.10, 0.20, 0.30, 0.40, 0.5] and u r = [0.20, 0.40, 0.60, 0.8]. The minimum ( u(1), u r ) works for all our worms was (0.08,0.20) –In practice, we use conservative value (0.15,0.5)

29 29 Results on Signature Quality (II) Suspicious pool with high noise ratio: –For noise ratio 50% and 70%, sometimes we can produce two signatures, one is the true worm signature, anther solely from noise. –The false positive of these noise signatures have to be very small: »Mean: 0.09% »Maximum: 0.7% Multiple worms with noises give similar results

30 30 Attack Resilience Assumptions Common assumptions for any sig generation sys 1.The attacker cannot control which worm samples are encountered by Hamsa 2.The attacker cannot control which worm samples encountered will be classified as worm samples by the flow classifier Unique assumptions for token-based schemes 1.The attacker cannot change the frequency of tokens in normal traffic 2.The attacker cannot control which normal samples encountered are classified as worm samples by the worm flow classifier

31 31 Normal Traffic Poisoning Attack We found our approach is not sensitive to the normal traffic pool used History: last 6 months time window The attacker has to poison the normal traffic 6 month ahead! 6 month the vulnerability may have been patched! Poisoning the popular protocol is very difficult.

32 32 Red Herring Attack Hard to implement Dynamic updating problem. Again our approach is fast Partial Signature matching, in extended version.

33 33 Coincidental Attack As mentioned in the Polygraph paper, increase the sample requirement Again, our approach are scalable and fast

34 34 Model Uniqueness of Invariants Let worm has a set of invariants: Determine their order by: t 1 : the token with minimum false positive in normal traffic. u(1) is the upper bound of the false positive of t 1 t 2 : the token with minimum joint false positive with t 1 FP({t 1,t 2 }) bounded by u(2) t i : the token with minimum joint false positive with {t 1, t 2, t i-1 }. FP({t 1,t 2,…,t i }) bounded by u(i) The total number of tokens bounded by k *

35 35 Problem Formulation Noisy Token Multiset Signature Generation Problem : INPUT : Suspicious pool  and normal traffic pool N; value  <1. OUTPUT : A multi-set of tokens signature S={(t 1, n 1 ),... (t k, n k )} such that the signature can maximize the coverage in the suspicious pool and the false positive in normal pool should less than  Without noise, exist polynomial time algo With noise, NP-Hard

36 36 Token-fit Attack Can Fail Polygraph Polygraph: hierarchical clustering to find signatures w/ smallest false positives With the token distribution of the noise in the suspicious pool, the attacker can make the worm samples more like noise traffic –Different worm samples encode different noise tokens Our approach can still work!

37 37 Token-fit attack could make Polygraph fail Noise samples N1 N2 N3 Worm samples W1 W2 W3 Merge Candidate 1 Merge Candidate 2 Merge Candidate 3 CANNOT merge further! NO true signature found!

38 38 Generalizing Signature Generation with noise BEST Signature = Balanced Signature –Balance the sensitivity with the specificity –But how? Create notation Scoring function: score(cov, fp, …) to evaluate the goodness of signature –Current used »Intuition: it is better to reduce the coverage 1/a if the false positive becomes 10 times smaller. »Add some weight to the length of signature (LEN) to break ties between the signatures with same coverage and false positive

39 39 Generalizing Signature Generation with noise Algorithm: similar Running time: same as previous simple form Attack Resilience Guarantee: similar

40 40 Extension to multiple worm Iteratively use single worm detector to detect multiple worm –At the first iteration, the algorithm find the signature for the most popular worms in the suspicious pool. All other worms and normal traffic treat as noise. –Though the analysis for the single worm can apply to multiple worms, but the bound are not very promising. Reason: high noise ratio

41 41 Implementation details Token Extraction: extract a set of tokens with minimum length l and minimum coverage COV min. –Polygraph use suffix tree based approach: 20n space and time consuming. –Our approach: Enhanced suffix array 8n space and much faster! (at least 20 times) Calculate false positive when check U-bounds –Again suffix array based approach, but for a 300MB normal pool, 1.2GB suffix array still large! –Optimization: using MMAP, memory usage: 150 ~ 250MB

42 42 Token Extraction Extract a set of tokens with minimum length l min and coverage COV min. And for each token output the frequency vector. Polygraph use suffix tree based approach: 20n space and time consuming. Our approach: –Enhanced suffix array 4n space –Much faster, at least 50(UPDATE) times! –Can apply to Polygraph also.

43 43 Calculate the false positive We need to have the false positive to check the U-bounds Again suffix array based approach, but for a 300MB normal pool, 1.2GB suffix array still large! Improvements –Caching –MMAP suffix array. True memory usage: 150 ~ 250MB. –2 level normal pool –Hardware based fast string matching –Compress normal pool and string matching algorithms directly over compressed strings

44 44 Future works Enhance the flow classifiers –Cluster suspicious flows by return messages –Malicious flow verification by replaying to Address Space Randomization enabled servers.

45 45 Experiment: Attacks We propose a new attack: token-fit. –The attacker may study the noise inside the suspicious pool –Create worm sample W i which may has more same tokens with some normal traffic noise sample N i –This will stuck the hierarchical clustering used in [Polygraph] –BUT We still can generate correct signature!


Download ppt "Hamsa: Fast Signature Generation for Zero-day Polymorphic Worms with Provable Attack Resilience Zhichun Li, Manan Sanghi, Yan Chen, Ming-Yang Kao and Brian."

Similar presentations


Ads by Google