Download presentation
Presentation is loading. Please wait.
Published byJonathan Cunningham Modified over 9 years ago
1
Coded Modulation and the Arrival of Signcryption Yuliang Zheng University of North Carolina at Charlotte yzheng@uncc.edu Enhancing Crypto-Primitives with Techniques from Coding Theory NATO Advanced Research Workshop 6-9 Oct. 2008, Veliko Tarnovo, Bulgaria
2
1 Outline Exposure to coded modulation Motivations Signcryption Look into the future
3
Error Corr (Encoder) Error Corr (Encoder) Modulation Error Corr (Decoder) Error Corr (Decoder) Security (Authen) Security (Authen) Security (Decryptor) Security (Decryptor) Security (Authen) Security (Authen) Security (Encryptor) Security (Encryptor) Source Decoder Source Decoder Source Encoder Source Encoder Communications system 2 Demodulation Channel
4
Prof. Hideki Imai’s Lab, in the late 80’s 3 Error Corr (Encoder) Error Corr (Encoder) Modulation Error Corr (Decoder) Error Corr (Decoder) Security (Authen) Security (Authen) Security (Decryptor) Security (Decryptor) Security (Authen) Security (Authen) Security (Encryptor) Security (Encryptor) Source Decoder Source Decoder Source Encoder Source Encoder Demodulation Channel
5
Error Correcting (encoder) + Modulation Error Correcting (encoder) + Modulation Error Correcting (decoder) + Demodulation Error Correcting (decoder) + Demodulation Coded modulation --- one of the hottest in 80’s 4 Security (Authen) Security (Authen) Security (Decryptor) Security (Decryptor) Security (Authen) Security (Authen) Security (Encryptor) Security (Encryptor) Source Decoder Source Decoder Source Encoder Source Encoder Channel
6
Coded modulation Combination/Co-design of error-correcting codes Multi-level modulation 2 types Trellis-coded modulation (TCM) Gottfried Ungerboeck, Proposed in1975; Published in IEEE Trans on IT in 1982 Block-coded modulation (BCM) Hideki Imai and Shuji Hirakawa Published in IEEE Trans on IT in 1976 5
7
Goals of coded modulation To transmit data as fast as possible, and as reliable (little errors) as possible To achieve Shannon’s capacity limit: C = B log 2 (1+S/N) C: channel capacity (bps) (= max. data transmission rate without error) B: channel bandwidth (Hz) S: power of signal (W) N: power of noise (W) 6
8
Challenges Modulation Transmission rate can be increased by Increase the number of waveforms (constellation points) for modulation However, an enlarged waveform set will also decrease “spacing” between waveforms, resulting in larger chances of errors at receiver Error-correcting codes Increasing error-correcting capabilities requires appending more parity bits/symbols, whereby reducing effective transmission rate 7
9
Address the challenges Co-design/integrate multi-level modulation and convolutional error-correcting code Assign waveforms to code words in such a way that maximizes Euclidean distance between the waveforms that are the most likely to be confused
10
Ungerboeck’s set partitioning (16QAM constellation mapper) 9
11
Error Corr (encoder) Error Corr (encoder) Modulation Trellis coded modulation The end result Increase transmission rate while at the same time, reduce errors at receiver KILL 2 BIRDS IN 1 STONE 10
12
Importance of coded modulation 11
13
Musing on coded modulation ½ dozen of my fellow PhD students working on coded modulation Doing Imai Lab’s ritual for 5 years weekly, long group seminars with students working on a vast array of different projects Applicable to cryptography ? Continue to muse after moving to Down Under 12
14
Major goals of cryptography Authenticity Trusted parties --- symmetric/private key authentication Untrusted parties --- asymmetric/public key authentication (digital signature, unforgeability) Confidentiality Symmetric/private key encryption Asymmetric/public key encryption Reduce cost/overhead Computation (over large integers) Expansion in length (=communication overhead) 13
15
In the paper & ink world: Signature-then-Seal 14 To achieve: authenticity (unforgeability & non-repudiation) To achieve: confidentiality
16
In the digital world: Signature-then-Encryption 1515/65 1. Add Signature Alice signs a message m using her secret key, i.e. creating sig on m. 2. Do Encryption Alice encrypts (m,sig) using AES with k. Alice creates another data so that Bob can recover k. (Typically, Alice encrypts k using Bob’s public key). msig m k m mod exp
17
Signature-then-Encryption (based on Discrete Logarithm) 16 encrypted using a private key cipher with k used by the receiver to reconstruct k m sig gxgx communication overhead EXP=3+2.17
18
Cost of Signature-then-Encryption 17
19
Questions on Efficiency Can we do better than “signature followed by encryption” ? For resource-constrained applications Wireless mobile devices Smart card applications Can we learn from other disciplines such as Communication engineering Imai-Hirakawa block coded modulation Ungerboeck trellis coded modulation 18
20
Co-design of digital signature and public key encryption ? 19 ? ? Security (Authen) Security (Authen) Security (Decryptor) Security (Decryptor) Security (Authen) Security (Authen) Security (Encryptor) Security (Encryptor) Source Decoder Source Decoder Source Encoder Source Encoder Error Correcting (encoder) + Modulation Error Correcting (encoder) + Modulation Error Correcting (decoder) + Demodulation Error Correcting (decoder) + Demodulation Channel
21
Goal: Signcryption (1996) To achieve both confidentiality, authenticity unforgeability & non-repudiation With a significantly smaller comp. & comm. Cost Cost (signcryption) << Cost (signature) + Cost (encryption) 20
22
How to co-design Focus on discrete logarithm (DL) based schemes: ElGamal type public key encryption DSS/Schnorr type digital signature Notice both use an “ephemeral public key”: g x Let them share the same ephemeral public key ! Hide it, as long as it can be recovered by the recipient ! 21
23
Signcryption -- public & secret parameters 22 Public to all p : a large prime q : a large prime factor of p-1 g : 0<g<p & with order q mod p Two 1-way hash functions: G: {0,1}*→ {0,1} L H: {0,1}* → Z q (E,D) : private-key encryption & decryption algorithms Alice’s keys : secret key : public key Bob’s keys : secret key : public key
24
Signcryption and Unsigncryption 23 Signcryption by Alice x ← {1,…,q-1} at random T = y b x mod p r = H(m, y a, y b, T) s = x/(r+x a ) mod q k=G(T) c=E k (m) Output (c, r, s) Signcryption by Alice x ← {1,…,q-1} at random T = y b x mod p r = H(m, y a, y b, T) s = x/(r+x a ) mod q k=G(T) c=E k (m) Output (c, r, s) Unsigncryption by Bob Compute k=G(T) m=D k (c) Check if r = H(m, y a, y b, T) Output m if yes, and Reject otherwise Unsigncryption by Bob Compute k=G(T) m=D k (c) Check if r = H(m, y a, y b, T) Output m if yes, and Reject otherwise m (c, r, s) (c, r, s) m
25
Cost of Signcryption (based on Discrete Logarithm) 24 m encrypted using a private key cipher with k communication overhead sig EXP=1+1.17
26
Why 1.17 exponentiations? 25 This can be done in a smart way, costing only 1.17 exponentiations on average ! D. Knuth, Seminumerical Algorithms, Vol. 2 of The Art of Computer Programming, 2 nd edition, Addison-Wesley, Exercise 27, Pages 465 & 637.
27
DL Signcryption v.s. sign- then-encrypt 26 |p|=|n| # of multiplications (the smaller the better)
28
DL Signcryption v.s. sign- then-encrypt 27 comm. Overhead ((# of bits, the smaller the better)
29
The end result 28 С един куршум - два заека (With one bullet - two rabbits)
30
Other developments Extensions: pairing, factorization, …… Add “bells and whistles” Multi-recipients, proxy, blind, threshold, ring, ID based, …… Authenticated encryption (Authencryption) Co-design of shared key authentication and encryption 29 Public or Private key AuthenticityNon- repudiation International standards SigncryptionPublic keyYes On the way AuthencryptionPrivate keyYesNoOn the way
31
Beyond coded modulation and signcryption ? There is no crypto equivalent of “Shannon’s capacity limit” Good ? Or Bad ? 30 Unsigncryption Source Decoder Source Decoder Source Encoder Source Encoder Signcryption Error Correcting (encoder) + Modulation Error Correcting (encoder) + Modulation Error Correcting (decoder) + Demodulation Error Correcting (decoder) + Demodulation Channel
32
31
33
32 Много благодаря ! Thank you !
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.