explicit construction of multiple access channel
play

Explicit Construction of Multiple Access Channel Resolvability Codes - PowerPoint PPT Presentation

Explicit Construction of Multiple Access Channel Resolvability Codes from Source Resolvability Codes Rumia Sultana R emi A. Chou Wichita State University 2020 IEEE International Symposium on Information Theory 1/29 Outline Introduction


  1. Explicit Construction of Multiple Access Channel Resolvability Codes from Source Resolvability Codes Rumia Sultana R´ emi A. Chou Wichita State University 2020 IEEE International Symposium on Information Theory 1/29

  2. Outline Introduction and Model • Coding Scheme • Coding Scheme Analysis • 2/29

  3. Channel Resolvability Problem (P2P) [Han-Verdú 93] Channel Resolvability Problem (P2P) [ Han-Verd´ u’93 ] S ~ Unif � ( ) N � 1: N X ~ q ENC X f ( ) S � i 1 N q q Y X | Y X | � 1: N 1: N Y Y p � q 1: N 1: N Y Y ? � � p q 1: N 1: N Y Y • ( ) denotes the uniform distribution over . � � • P2P channel ( X , q Y | X , Y ) ( � , , � ) • P2P Channel . • Unif ( S ) denotes the uniform distribution over S | 3/29 3

  4. Channel Resolvability (MAC) [ Steinberg’98 ] Channel Resolvability (MAC) [Steinberg’98] S ~ Unif � ฀ ( ) S ~ Unif � ฀ ( ) 1 1 2 2 N � ENC ENC 1: N 1: N ( X , Y ) ~ q XY f ( S ) f ( S ) � i 1 1, N 1 2, N 2 q q Z XY | Z XY | � 1: N 1: N Z Z � q p 1: N 1: N Z Z ? � p � q 1: N 1: N z z • and are uniformly distributed sequences. • DMMAC ( X × Y , q Z | XY , Z ) 1 2 � ( , , ) � � � • DMMAC . • S 1 and S 2 are uniformly distributed sequences | 4 4/29

  5. Applications • Strong secrecy MAC-WT [ Pierrot-Bloch’11 ], [ Yassaee-Aref’10 ] • Cooperative jamming [ Pierrot-Bloch’11 ] • Semantic security for MAC-WT [ Frey-Bjelakovic-Stanczak’17 ] • Strong coordination in networks [ Bloch-Kliewer’13 ] 5/29

  6. Related Works (Non-exhaustive) P2P Resolvability • Existence results ⇒ [ Han-Verdu’93 ] • Explicit and low-complexity code constructions ⇒ Polar codes for resolvability over BSC and DMC [ Bloch-Luzzi-Kliewer’12 ], [ Chou-Bloch-Kliewer’18 ] ⇒ Injective group homomorphisms for resolvability over BSC [ Hayashi-Matsumoto’16 ] • Non-explicit, linear coding ⇒ Resolvability over DMC [ Amjad-Kramer’15 ] 6/29

  7. Related Works MAC Resolvability • Existence results ⇒ MAC [ Steinberg’98 ] ⇒ MAC resolvability region [ Frey-Bjelakovic-Stanczak’17 ] • Explicit low-complexity coding scheme ⇒ Symmetric MAC [ Chou-Bloch-Kliewer’14 ] − Invertible extractors − Injective group homomorphisms • Explicit low-complexity coding scheme ⇒ Polar codes for arbitrary MAC [ Sultana-Chou’19 ] 7/29

  8. This Work • Explicit and low-complexity codes for MAC resolvability ⇒ Arbitrary DMMAC ⇒ Input alphabets are binary • Coding scheme ideas ⇒ Reduce MAC resolvability coding problem to a simpler source resolvability coding problem ⇒ Randomness recycling with distributed hashing coupled with block-Markov encoding ⇒ Source resolvability codes used in a black box manner 8/29

  9. Outline Introduction and Model • Coding Scheme • Coding Scheme Analysis • 9/29

  10. Reduction of the MAC Resolvability Region to a Simpler Region To achieve the MAC resolvability region R q Z , it is sufficient to achieve regions of the form R X , Y for some fixed distribution p X p Y     ( R 1 , R 2 ) : I ( XY ; Z ) < R 1 + R 2 ,     R X , Y � I ( X ; Z ) < R 1 ,       I ( Y ; Z ) < R 2 10/29

  11. <latexit sha1_base64="trq93S7UP39rPUZHw8RvOGPbU=">AB7nicbVBNSwMxEJ31s9avqkcvwSJUkLJbBT0Wveitgv3QdinZNuGZrMhyQpl6Y/w4kERr/4eb/4b03YP2vpg4PHeDPzAsmZNq7Swtr6yurec28ptb2zu7hb39ho4TRWidxDxWrQBrypmgdcMpy2pKI4CTpvB8HriN5+o0iwW92YkqR/hvmAhI9hYqXlbaj2cPp50C0W37E6BFomXkSJkqHULX51eTJKICkM41rtudL4KVaGEU7H+U6iqcRkiPu0banAEdV+Oj13jI6t0kNhrGwJg6bq74kUR1qPosB2RtgM9Lw3Ef/z2okJL/2UCZkYKshsUZhwZGI0+R31mKLE8JElmChmb0VkgBUmxiaUtyF48y8vkal7J2VK3fnxepVFkcODuEISuDBVThBmpQBwJDeIZXeHOk8+K8Ox+z1iUnmzmAP3A+fwDVqY6V</latexit> <latexit sha1_base64="QhE/VZAMEGYPMCTnY4ZydI0eRE=">AB8HicbVA9TwJBEJ3zE/ELtbTZCbYkDsJLEhsdEOE/kSLmRv2YMNu3uX3T0TQvgVNhYaY+vPsfPfuMAVCr5kpf3ZjIzL4g508Z1v5219Y3Nre3MTnZ3b/gMHd03NBRogitk4hHqhVgTmTtG6Y4bQVK4pFwGkzGN3M/OYTVZpF8sGMY+oLPJAsZAQbK7ULd8VW+/rxotDL5d2SOwdaJV5K8pCi1st9dfsRSQSVhnCsdcdzY+NPsDKMcDrNdhNY0xGeEA7lkosqPYn84On6NwqfRGypY0aK7+nphgofVYBLZTYDPUy95M/M/rJCas+BMm48RQSRaLwoQjE6HZ96jPFCWGjy3BRDF7KyJDrDAxNqOsDcFbfnmVNMol7JUvi/nq5U0jgycwhkUwYMrqMIt1KAOBAQ8wyu8Ocp5cd6dj0XrmpPOnMAfOJ8/ovKO9A=</latexit> <latexit sha1_base64="FWHmvPE/De9xswNawOfWxso7D6E=">AB6nicbVA9TwJBEJ3DL8Qv1NJmI5hYkTspCShscQoHwlcyN6ywIa9vcvunAm58BNsLDTG1l9k579xgSsUfMkL+/NZGZeEth0HW/ndzW9s7uXn6/cHB4dHxSPD1rmyjRjLdYJCPdDajhUijeQoGSd2PNaRhI3gmjYXfeLaiEg94izmfkjHSowEo2ilh3KjPCiW3Iq7BNkXkZKkKE5KH71hxFLQq6QSWpMz3Nj9FOqUTDJ54V+YnhM2ZSOec9SRUNu/HR56pxcWVIRpG2pZAs1d8TKQ2NmYWB7QwpTsy6txD/83oJjmp+KlScIFdstWiUSIRWfxNhkJzhnJmCWVa2FsJm1BNGdp0CjYEb/3lTdKuVrybSvW+WqrXsjycAGXcA0e3EId7qAJLWAwhmd4hTdHOi/Ou/Oxas052cw5/IHz+QNLFY0b</latexit> <latexit sha1_base64="FWHmvPE/De9xswNawOfWxso7D6E=">AB6nicbVA9TwJBEJ3DL8Qv1NJmI5hYkTspCShscQoHwlcyN6ywIa9vcvunAm58BNsLDTG1l9k579xgSsUfMkL+/NZGZeEth0HW/ndzW9s7uXn6/cHB4dHxSPD1rmyjRjLdYJCPdDajhUijeQoGSd2PNaRhI3gmjYXfeLaiEg94izmfkjHSowEo2ilh3KjPCiW3Iq7BNkXkZKkKE5KH71hxFLQq6QSWpMz3Nj9FOqUTDJ54V+YnhM2ZSOec9SRUNu/HR56pxcWVIRpG2pZAs1d8TKQ2NmYWB7QwpTsy6txD/83oJjmp+KlScIFdstWiUSIRWfxNhkJzhnJmCWVa2FsJm1BNGdp0CjYEb/3lTdKuVrybSvW+WqrXsjycAGXcA0e3EId7qAJLWAwhmd4hTdHOi/Ou/Oxas052cw5/IHz+QNLFY0b</latexit> <latexit sha1_base64="me5at2af0sjJOyZEIuwvWlpX0WQ=">AB73icbVA9TwJBEJ3DL8Qv1NJmI5hgQ+6gkMSGxEY7TORD4UL2lj3YsLd37u6ZkAt/wsZCY2z9O3b+Gxe4QsGXTPLy3kxm5nkRZ0rb9reVWVvf2NzKbud2dvf2D/KHRy0VxpLQJgl5KDseVpQzQZuaU47kaQ48Dhte+Ormd9+olKxUNzpSUTdA8F8xnB2kid4k3p/vLhvNjPF+yPQdaJU5KCpCi0c9/9QYhiQMqNOFYqa5jR9pNsNSMcDrN9WJFI0zGeEi7hgocUOUm83un6MwoA+SH0pTQaK7+nkhwoNQk8ExngPVILXsz8T+vG2u/5iZMRLGmgiwW+TFHOkSz59GASUo0nxiCiWTmVkRGWGKiTUQ5E4Kz/PIqaVXKTrVcua0U6rU0jiycwCmUwIELqM1NKAJBDg8wyu8WY/Wi/VufSxaM1Y6cwx/YH3+APZEjpI=</latexit> <latexit sha1_base64="guWHrgPXQvE6tkAEcBkQJWPBvA=">AB73icbVBNT8JAEJ3iF+IX6tHLRjDBC2nxIkXEi96w0SgERqyXbawYbutu1sT0vAnvHjQGK/+HW/+GxfoQcGXTPLy3kxm5vkxZ0rb9reVW1vf2NzKbxd2dvf2D4qHR20VJZLQFol4JF0fK8qZoC3NKduLCkOfU47/vh65neqFQsEvd6ElMvxEPBAkawNpJbvq24Vw/n5X6xZFftOdAqcTJSgzNfvGrN4hIElKhCcdKdR071l6KpWaE02mhlygaYzLGQ9o1VOCQKi+d3ztFZ0YZoCSpoRGc/X3RIpDpSahbzpDrEdq2ZuJ/3ndRAd1L2UiTjQVZLEoSDjSEZo9jwZMUqL5xBMJDO3IjLCEhNtIiqYEJzl1dJu1Z1Lq1u1qpUc/iyMJnEIFHLiEBtxAE1pAgMzvMKb9Wi9WO/Wx6I1Z2Uzx/AH1ucP9LyOkQ=</latexit> <latexit sha1_base64="oHnBRhRnldpNoE98AjF3zW4mkw=">AB7HicbVA9TwJBEJ3DL8Qv1NJmI5hYkTspCSxsUTjIQlcyN6ywIa9vcvunAm58BtsLDTG1h9k579xgSsUfMkL+/NZGZemEh0HW/ncLG5tb2TnG3tLd/cHhUPj5pmzjVjPslrHuhNRwKRT3UaDknURzGoWSP4aTm7n/+MS1EbF6wGnCg4iOlBgKRtFKfvW+71X75Ypbcxcg68TLSQVytPrlr94gZmnEFTJjel6boJBRjUKJvms1EsNTyib0BHvWqpoxE2QLY6dkQurDMgw1rYUkoX6eyKjkTHTKLSdEcWxWfXm4n9eN8VhI8iESlLki0XDVNJMCbz8lAaM5QTi2hTAt7K2FjqilDm0/JhuCtvrxO2vWad1Wr39UrzUYeRxHO4BwuwYNraMItMAHBgKe4RXeHOW8O/Ox7K14OQzp/AHzucPh9yNzg=</latexit> <latexit sha1_base64="3AGOXShm9AD17XUCLDiflnOLeag=">AB7HicbVBNT8JAEJ3iF+IX6tHLRjDxRNp6kCOJF49oLJAQ7bLFjbsbpvdrQlp+A1ePGiMV3+QN/+NC/Sg4EsmeXlvJjPzopQzbVz32yltbG5t75R3K3v7B4dH1eOTjk4yRWhAEp6oboQ15UzSwDaTdVFIuI08docjP3H5+o0iyRD2a0lDgkWQxI9hYKajfD/z6oFpzG+4CaJ14BalBgfag+tUfJiQTVBrCsdY9z01NmGNlGOF0VulnmqaYTPCI9iyVWFAd5otjZ+jCKkMUJ8qWNGih/p7IsdB6KiLbKbAZ61VvLv7n9TITN8OcyTQzVJLlojyCRo/jkaMkWJ4VNLMFHM3orIGCtMjM2nYkPwVl9eJx2/4V01/Du/1moWcZThDM7hEjy4hbcQhsCIMDgGV7hzZHOi/PufCxbS04xcwp/4Hz+AIlhjc8=</latexit> Reduction of the MAC Resolvability Region to a Simpler Region Case 1. I ( XY ; Z ) = I ( X ; Z ) + I ( Y ; Z ) R 2 R X,Y <latexit sha1_base64="u+G/yn4iQeq3wbKGVlO9NC5NRos=">AB+nicbVDLSsNAFL3xWesr1aWbwSK4kJUQZdFNy6r2Ie0IUym03boZBJmJkqJ/RQ3LhRx65e482+ctFlo64GBwzn3cs+cIOZMacf5tpaWV1bX1gsbxc2t7Z1du7TXVFEiCW2QiEeyHWBFORO0oZnmtB1LisOA01Ywusr81gOVikXiTo9j6oV4IFifEayN5Nulboj1kGCObv20fXI/Kfp2ak4U6BF4uakDnqv3V7UkCanQhGOlOq4Tay/FUjPC6aTYTRSNMRnhAe0YKnBIlZdOo0/QkVF6qB9J84RGU/X3RopDpcZhYCazoGrey8T/vE6i+xdeykScaCrI7FA/4UhHKOsB9ZikRPOxIZhIZrIiMsQSE23aykpw57+8SJrVintaqd6clWuXeR0FOIBDOAYXzqEG1CHBhB4hGd4hTfryXqx3q2P2eiSle/swx9Ynz8dFJM/</latexit> C I ( XY, Z ) C I ( Y ; Z ) R 1 I ( X ; Z ) I ( XY ; Z ) 11/29

Recommend


More recommend