Spatially Coupled LDPC Codes

Size: px
Start display at page:

Download "Spatially Coupled LDPC Codes"

Transcription

1 Spatially Coupled LDPC Codes Kenta Kasai Tokyo Institute of Technology 30 Aug, 2013

2 We already have very good codes. Efficiently-decodable asymptotically capacity-approaching codes Irregular LDPC Codes Achievability: proof is limited to BEC but likely achieve any memoryless channels. Drawback: non-universal and high error floors Polar Codes Achievability: proved for any memoryless channels Drawback: poor finite length performance

3 Optimal Finite-Length Error-Rate Approximation log M (n, P B ) = nc(ϵ) nv (ϵ)q 1 (P B ) + O(log n) ( n(c(ϵ) ) R ) P B = Q (1 + o(1)) V (ϵ) R := 1 n log M (n, P B ) M (n, P B ) is the maximum number of codewords of length-n codes that achieve the block error rate P B ϵ : channel parameter C(ϵ) : channel capacity V (ϵ) : channel dispersion, e.g., V (ϵ) = ϵ(1 ϵ) + log 2 ( ) 1 ϵ ϵ for BSC(ϵ) V (ϵ) = ϵ(1 ϵ) for BEC(ϵ).

4 Example: Fix BEC(ϵ=0.5) and BSC(ϵ=0.11), C(ϵ) = 0.5 and P B = Who wants to fill the gap 1% away from capacity? n = 10 5 is sufficient.

5 Example: State-of-the-art finite-length codes. Fix R and P B over BEC(ϵ) Channel Capacity C(ε)=1-ε Finite Polar Code Infinite (2,4)-NBLDPCC GF(2 8 ) Finite (2,4)-NBLDPCC GF(2 8 ) Infinite (3,6)-BLDPCC Finite (3,6)-BLDPCC Finite Optimal Code Block Length n [bit]

6 Uncoupled LDPC Code Sparse parity-check matrix (column-weight 2, row-weight 4) H = , C = {x Fn 2 Hx T = 0} Sparse graph representation (variable-degree 2, check-degree 4)

7 Efficient locally-optimal decoding uncoupled LDPC code x 5 + x 7 + x 10 + x 12 = 0 ( ) x 5 x 7 x 10 x 12 Suppose we observed x 5 =?, x 7 = 1, x 10 = 1, x 12 = 0 through the BEC(ϵ), then from ( ) we know x 5 = 1. Iterate this until decoding is stuck. Locally-optimal decoding is efficient with computational cost O(n).

8 Locally-optimal decoder performance analysis of (d l, d r )-codes over BEC(ϵ) Erasure probability of check node of degree d r q (t) = 1 (1 p (t 1) ) d r 1 Erasure probability of variable node of degree d l Putting together p (t) = ϵ(q (t) ) d l 1 p (t) = ϵ(1 (1 p (t 1) ) d r 1 ) d l 1 = f(g(p (t 1) ); ϵ) How good the (d l, d r )-code is measured by threshold ϵ BP (d l, d r ) = sup{ϵ lim t p (t) = 0}

9 Globally optimal decoder performance analysis x : codeword of (l, r) code y : received output of channel ˆx MAP i (y) = argmax P (x i y) x i F 2 { ϵ MAP = sup ϵ 1 n n Pr(X i i=1 } MAP ˆX i (Y )) = 0 Globally optimal decoding is computationally complex with cost O(2 nr ).

10 Graphical interpretation of ϵ BP (d l, d r ) via potential U(x; ϵ) Potential function U(x; ϵ) = xg(x) G(x) F (g(x); ϵ) ϵ BP (d l, d r ) = sup{ϵ U (x; ϵ) > 0 x(0, 1]} ϵ POT BEC (d l, d r ) := sup{ϵ min U(x; ϵ) > 0} = ϵ MAP (d l, d r ) x [0,1] Example: (d l = 3, d r = 6) uncoupled code. ϵ BP (d l, d r ) ϵ POT (d l, d r ) Reprinted from [Yedla et al.]

11 How to couple L uncoupled (d l, d r ) codes to make a (d l, d r, L) coupled code 1. Copy L (d l, d r ) uncoupled codes L 2. Re-connect edges randomly with w neighboring codes

12 Locally-optimal decoder performance analysis of (d l, d r, L)-coupled codes over BEC(ϵ) x i : check node erasure probability of i-th uncoupled code. DEMO. vector form: x (t+1) i = 1 w w 1 k=0 ( w 1 1 f w j=0 ) g(x (t) i+j k ); ϵ i k x (t+1) = A T f(ag(x (t) ); ϵ) (A is some upper-triangle matrix)

13 (d l, d r, L, w) coupled codes achieve ϵ POT (d l, d r ) Definition: potential function of (d l, d r, L, w) coupled codes U(x; ϵ) = g(x) T x G(x) F (Ag(x); ϵ) Theorem [Yedla et al., 2012] An x [0, 1] L is a stationary point of U(x; ϵ) if and only if x is a fixed point of BP iteration x = A T f(ag(x); ϵ). For ϵ < ϵ POT (d l, d r ), if w is large enough, the only stationary point of U(x; ϵ) is 0.

14 History of Coupled LDPC Codes Felström and Zigangirov constructed coupled codes as LDPC convoluional codes and found they were better than uncoupled codes. Lentmaier found (4, 8, L) coupled codes had low error rate at ϵ above the BP threshold (4, 8) uncoupled codes. Sridharan et al. found the BP threshold of coupled codes are very close to the MAP threshold of the uncoupled codes.

15 History of Coupled LDPC Codes (Cont.) Kudekar et al. proved lim w ϵbp (d l, d r, L) = ϵ MAP (d l, d r ). Yedla et al. proved lim w ϵbp (d l, d r, L, w) = ϵ MAP (d l, d r ) = ϵ POT (d l, d r ). Kudekar et al. reported empirical evidence that this phenomenon also occurs for MBIOS channels other than BEC. Kudekar et al. reported empirical evidence that this phenomenon occurs for channels with memory and multi-access channels.

16 Practical Definition of Coupled Codes via Protograph Protograph A protograph is a bipartite graph defined by a set of variable nodes V = {v 1,..., v n } a set of check nodes C = {c 1,..., c m },and a set of edges E. Base matrix A protograph is defined by a base matrix B F m n 2. v 1 v 2 v 3 v 4 v 5 v 6 c c c c

17 Protograph Codes A protograph code is defined by a parity-check matrix H F mm nm 2. C := {x F nm 2 Hx T = 0} a parity-check matrix H is derived from a base matrix B. 1 = M M size random binary permutation matrix. 0 = M M size zero matrix. The parameter M is referred to as lifting number. v 1 v 2 v 3 v 4 v 5 v 6 c 1 P 1 P 2 P c 2 0 P 4 P 5 P c 3 P P 8 P 9 c P 10 P 11 P 12

18 (d l, d r, L) coupled codes The base matrix B(d l, d r, L) is a (L + d l 1) kl binary band matrix of band size d r d l. B(d l = 4, d r = 12, L = 7) =

19 ... Ex. parity-check matrix for (d l =4, d r = 12, L = 7).. P 1,10P 1,11P 1,12 P 2,7 P 2,8 P 2,9 P 2,10P 2,11P 2,12 P 3,4 P 3,5 P 3,6 P 3,7 P 3,8 P 3,9 P 3,10P 3,11P 3,12 P 4,1 P 4,2 P 4,3 P 4,4 P 4,5 P 4,6 P 4,7 P 4,8 P 4,9 P 4,10P 4,11P 4,12 P 5,1 P 5,2 P 5,3 P 5,4 P 5,5 P 5,6 P 5,7 P 5,8 P 5,9 P 5,10P 5,11P 5,12 P 6,1 P 6,2 P 6,3 P 6,4 P 6,5 P 6,6 P 6,7 P 6,8 P 6,9 P 6,10P 6,11P 6,12 P 7,1 P 7,2 P 7,3 P 7,4 P 7,5 P 7,6 P 7,7 P 7,8 P 7,9 P 7,10P 7,11P 7,12 P 8,1 P 8,2 P 8,3 P 8,4 P 8,5 P 8,6 P 8,7 P 8,8 P 8,9 P 9,1 P 9,2 P 9,3 P 9,4 P 9,5 P 9,6 P 10,1P 10,2P 10,3 Denote a codeword by (x 1,...,x 21 ). The parity-check equations are written by dr j=1 P i,j x ik dr+j =0 (for i =0,...,10) (1).. P i,j s are M M random permutation matrices. M 1000 Rate of (d l, d r ) uncoupled codes R(d l, d r ) = k 1 k, k := d r d l Rate of (d l, d r, L) coupled code R(d l, d r, L) = k 1 k d l 1 kl

20 .. Encoding of Coupled Codes. Ex. parity-check matrix for (d l =4, d r = 12, L = 7).. P 1,10P 1,11P 1,12 P 2,7 P 2,8 P 2,9 P 2,10P 2,11P 2,12 P 3,4 P 3,5 P 3,6 P 3,7 P 3,8 P 3,9 P 3,10P 3,11P 3,12 P 4,1 P 4,2 P 4,3 P 4,4 P 4,5 P 4,6 P 4,7 P 4,8 P 4,9 P 4,10P 4,11P 4,12 P 5,1 P 5,2 P 5,3 P 5,4 P 5,5 P 5,6 P 5,7 P 5,8 P 5,9 P 5,10P 5,11P 5,12 P 6,1 P 6,2 P 6,3 P 6,4 P 6,5 P 6,6 P 6,7 P 6,8 P 6,9 P 6,10P 6,11P 6,12 P 7,1 P 7,2 P 7,3 P 7,4 P 7,5 P 7,6 P 7,7 P 7,8 P 7,9 P 7,10P 7,11P 7,12 P 8,1 P 8,2 P 8,3 P 8,4 P 8,5 P 8,6 P 8,7 P 8,8 P 8,9 P 9,1 P 9,2 P 9,3 P 9,4 P 9,5 P 9,6 P 10,1P 10,2P 10,3 Denote a codeword by (x 1,...,x 21 ). The parity-check equations are written by dr j=1 P i,j x ik dr+j =0 (for i =0,...,10) (1).. The computational cost of the sequential encoding is O(LM). While the computational cost of the termination is O(M 2 ).

21 Termination Termination calculates values of parity bits x 17, x 18, x 19, x 20, x 21 so that constraints c 6, c 7, c 8, c 9, c 10 are satisfied. The termination is equivalent to solve this equation. P 6,11 P 6,12 x 17 s 6 P 7,8 P 7,9 P 7,10 P 7,11 P 7,12 x 18 P 8,5 P 8,6 P 8,7 P 8,8 P 8,9 x 19 P 9,2 P 9,3 P 9,4 P 9,5 P 9,6 x 20 = s 7 s 8 s 9 P 10,1 P 10,2 P 10,3 To solve this equation involves multiplication of 5M 5M size not sparse matrix, so one needs O(M 2 ) computational cost to finish the termination. s i denote syndrome of c i which contains information of previously decided bits. x 21 s 10

22 ... Modified (d l, d r, L) codes.. Modified (d l, d r, L) codes is protograph codes. The protograph of modified (d l, d r, L) codes is obtained by removing d l 2check nodes of the protograph of (d l, d r, L) codes. kl (L + 1) The design rate is = k kl k kl Parity-check matrix of modified (4,12,7) codes P 1,10P 1,11P 1,12 P 2,7 P 2,8 P 2,9 P 2,10P 2,11P 2,12 P 3,4 P 3,5 P 3,6 P 3,7 P 3,8 P 3,9 P 3,10P 3,11P 3,12 P 4,1 P 4,2 P 4,3 P 4,4 P 4,5 P 4,6 P 4,7 P 4,8 P 4,9 P 4,10P 4,11P 4,12 P 5,1 P 5,2 P 5,3 P 5,4 P 5,5 P 5,6 P 5,7 P 5,8 P 5,9 P 5,10P 5,11P 5,12 P 6,1 P 6,2 P 6,3 P 6,4 P 6,5 P 6,6 P 6,7 P 6,8 P 6,9 P 6,10P 6,11P 6,12 P 7,1 P 7,2 P 7,3 P 7,4 P 7,5 P 7,6 P 7,7 P 7,8 P 7,9 P 7,10P 7,11P 7,12 P 8,1 P 8,2 P 8,3 P 8,4 P 8,5 P 8,6 P 8,7 P 8,8 P 8,9. The modified coupled codes still have the same BP threshold as the unmodified coupled codes.

23 Sliding Window Decoding Performance Approaches Full Decoding Performance Theorem [Iyengar et al., 2012] Consider windowed decoding of the (d l, d r, w, L) spatially coupled ensemble over the binary erasure channel. Then for a target erasure rate δ < δ, there exists a positive integer W min (δ) such that when the window size W W min (δ) the WD threshold satisfies ( ϵ WD (d l, d r, w, W, δ) 1 d ) ld r 2 δ dl 2 d l 1 ( ϵ BP (d l, d r, L =, w) e 1 B ( W w 1 A ln ln D δ C)). Here A, B, C, D and δ are strictly positive constants that depend only on the ensemble parameters d l, d r and w. This theorem says that the WD thresholds approach the BP threshold ϵ BP (d l, d r, L =, w) at least exponentially fast in the ratio of the size of the window W to the coupling length w for a fixed target erasure probability δ < δ.

24 Multi Dimensional Extension Background Spatially-Coupled (SC) codes are constructed by coupling many regular low-density parity-check codes in a chain. Strongness: Spatial coupling improves BP threshold. Drawback: The decoding chain of SC codes aborts when facing burst erasure. Goal In this paper, we introduce multi-dimensional (MD) SC codes to circumvent the burst erasures.

25 Definition: 1D-SC (d l, d r, L, w, Z) Codes Consider L sections on 1D discrete torus Z L. Put M bit nodes of degree d l and d l d r M check nodes of degree d r at each section i {0,..., L 1}. Connect edges randomly between bit nodes at section i and check nodes at section i + j (j {0,..., w 1}) with Md l /w edges. Shorten all the bit nodes at section i Z = {0,..., w 1} Figure : (d l = 3, d r = 6, L = 5, w = 2, Z = {0, 1}) code

26 We Introduce Simple Representation of Connection in Order to Emphasize Symmetry of Decoding Messages w = 2 w = 3 In right figure, nodes are connected if the corresponding sections can be reached within 2 edges in left figure.

27 Definition: (d l, d r, L, w, Z) 2D-SC Code Consider L 2 sections on 2D-torus Z 2 L. Put M bit nodes of degree d l and d l d r M check nodes of degree d r at each section i {0,..., L 1} 2. Connect randomly bit nodes at section i and check nodes at section i + j (j {0,..., w 1} 2 ) with Md l /w 2 edges. Shorten all the bit nodes at section i Z Z 2 L. Example Figure : w = 2 Figure : w = 3

28 Density Evolution Predicts BP Erasure Probability p (l) i round l at section i at p (0) i = p (l) i = p (0) i 0 (i Z) ϵ (i / Z) ϵ + α ( burst erasures occur at section i) ( j 1 ( w D 1 (1 k 1 ) ) w D p(l 1) i+j k ) dr 1 d l 1 We are interested in BP threshold { ϵ such that } BP threshold: ϵ = sup ϵ > 0 lim P (l) b = 0 l BP erasure probability: P (l) b = 1 L D p d l i #Z i Z D L

29 2D-SC Codes with Line Segment Zero Domain are Equivalent to 1D-SC Codes Proposition: ϵ 2D(d l, d r, L, w, Z 2D ) = ϵ 1D(d l, d r, L, w, Z 1D ) { } Z 2D = (i, j) i {0, 1,..., w 1}, j [L] Z 1D = {0, 1,..., w 1} Example Left figure shows initial constellation p (0) of 2D-SC codes. (i 1,i 2 ) There are L 2 = sections Black line at middle represents zero domain Z Red sections are set with ϵ = Demonstrate the density evolution of (3,6) 2D-SC codes

30 Burst Error is a Problem for SC Codes SC codes have a nice structure for windowed encoding and decoding. Interleaving over whole coded bits kills the structure. Burst errors stops the decoding chain. Increasing coupling number L does not help. Increasing randomized window size w does help. But large w is not preferable. We assume interleaver in each section. Then burst erasures can be viewed as output from BEC(ϵ + α). We assume burst erasures occur randomly at some O(1) sections.

31 i 1D-SC decoding is a chain reaction like dominoes. If one of dominoes is too heavy to be toppled, the decoding chain reaction aborts. Example: (d l = 3, d r = 6, L = 101, w = 4, Z = { 1, 0, +1}) BEC(ϵ = 0.48) + burst erasures BEC(ϵ + α = 0.52) Erasure Probability p (l) iteration= 0 iteration= 50 iteration= 100 iteration= 300 iteration= 700 iteration= Section i Example:

32 2D-SC Codes are More Robust to Burst Erasures than 1D-SC Codes 0.5 BP threshold ϵ Uncoupled BP threshold Uncoupled MAP threshold 0.1 1D with 1 burst section erasure 1D with 2 burst section erasures 2D with 1 burst section erasure 2D with 2 burst section erasures Number of coupled neighboring sections: Adjacent Section Number w or w 2 Theorem: Necessity Condition of Decoding Burst Erasure The MD (d l, d r, L, ω, Z) code of dimension D cannot recover any burst section erasure if ϵ + α > ϵ BP (d l, d r )w D, where ϵ BP (d l, d r ) is the BP threshold of uncoupled (d l, d r ) codes.

33 Square Zero Domain Mitigates Rate-loss 2D-SC codes have L 2 sections. But 2D-SC codes with line segment zero domain have rate-loss O(1/L) which is worse than 1D-SC codes. Rate-loss of 2D-SC codes with square zero domain with size z turns out O(1/L 2 ) Demonstrate density evolution Square zero domain size z has to be set large to grow. Zero domain surface needs to be flat to grow in the discrete world. Small size zero domain degrades BP threshold.

34 Square Zero Domain Mitigates Rate-loss BP threshold ϵ Uncoupled MAP threshold 2D Coupled BP Threshold z

35 Conclusion We proposed multi-dimensional SC codes. 2D-SC codes achieve the same BP threshold as 1D-SC codes when line-segment zero domains are used as the shortened domain. 2D-SC codes are more robust against burst section erasures than 1D-SC codes.

36 Good points of spatially-coupled LDPC codes Minimum distance and water-fall tradeoff Irregular LDPC codes often suffer from the dilemma between minimum distance and water-fall performance. Spatially-coupled LDPC codes have, at the same time, large minimum distance and capacity-achieving water-fall performance. Universality Sason et al. proved that, in the limit of large column-weight and row-weight, regular LDPC codes achieve the capacity of any MBIOS channels under MAP decoding. lim d l,d r ϵmap (d l, d r ) = ϵ Shannon (d l, d r ) Hence, it is expected that coupled codes universally achieve the capacity.

37 Drawback of coupled codes Drawback Infinite column-weight and row-weight are required to strictly achieve the capacity, which leads to infinite computation per iteration. Bad news Sason and Urbanke showed the codes achieving 1 ϵ of capacity under MAP decoding have average degree at least O(ln 1 ϵ ) This result is valid only for the codes without punctured bits.

38 Capacity-Achieving Codes with Bounded Degree MacKay-Neal (MN) Codes Murayama et al. Tanaka and Saad observed by replica method that MN codes were capacity-achieving under MAP decoding with bounded degree for BSC and AWGNC. Hsu-Anastasopoulos (HA) Codes Hsu and Anastasopoulos proved HA codes achieve the capacity of any MBIOS channels under MAP decoding with bounded degree.

39 Couped MN codes achieve Shannon limit For any d l, d r, d g 2, we observed the BP thresholds of spatially coupled (d l, d r, d g )-MN codes are very close to the Shannon limit for the BEC. Obata et al. proved that ϵ POT MN (d l = 2, d r, d g = 2) = ϵ Shannon MN (d l = 2, d r, d g = 2). Protograph of spatially coupled (8,4,5)-MN code

40 Couped HA codes achieve Shannon limit For any d l, d r, d g 2, we observed the BP thresholds of spatially coupled (d l, d r, d g )-HA codes are very close to the Shannon limit for the BEC. Obata et al. proved that ϵ POT HA (d l, d r = 2, d g = 2) = ϵ Shannon HA (d l, d r = 2, d g = 2). Protograph of spatially coupled (4,8,5)-HA code

41 Uncoupled Codes and How to Couple Them (d l, d r )-regular codes x d r x d l Density evolution updates scalers x (l) R, (l = 0, 1,... ) (d l, d r, L, w)-coupled codes Shortened {}}{ 0 {}}{ L 1 {}}{ Shortened {}}{ x d r x d l Density evolution updates vectors x (l) R L, (l = 0, 1,... )

42 (d l, d r, d g, L, w) Coupled Hsu-Anastasopoulos (HA) Codes Achieve Capacity with BP Decoding with Bounded Degree Achieve the capacity of BEC with bounded degree (proved). Numerical results show universality over BMS channels, e.g., BAWGNC [Mitchell et al., 12]. x d r Shortened {}}{ 0 {}}{ L 1 {}}{ Shortened {}}{ x d l x d g x d g

43 Block Codes and Rateless Codes Block codes Block codes are designed to generate n fixed-length coded bits from k information bits so that the coding rate R = k n approaches the capacity C with vanishing decoding error probability. Rateless Codes Rateless codes are designed to generate limitless sequence of coded bits from k information bits so that receivers can recover the k information bits from arbitrary (1 + α)k/c =: n received bits with vanishing overhead α and with vanishing decoding error probability. Two preceding rateless codes 1. Raptor Codes 2. Coupled Rateless LDGM Codes

44 Raptor Codes Have Unbounded Degree Scheme [Shokrollahi, 03] High-rate regular LDPC pre-code Each output bit is the sum of d randomly selected pre-coded bits w.p. Ω d Optimized irregular output degree distribution Ω(x) = d Ω dx d What is strong Raptor codes achieve capacity of BEC(ϵ). Drawback x d r x d l e l avg(x 1) Ω(x) Raptor codes need to have unbounded maximum degree d of Ω(x) to achieve capacity, which leads to a computation complexity problem both at encoders and decoders. Being optimized for only BEC, raptor codes are not universal for BMSC.

45 Coupled Rateless LDGM Codes Have Non-zero Error Probability Shortened 0 L 1 {}}{{}}{{}}{ Shortened {}}{ e l avg(x 1) x d Scheme [Aref and Urbanke, 11] Coupled rateless LDGM code No precode What is strong Achieves area threshold of the underlying code Coupled rateless LDGM codes are likely universal Drawback Degree 0 information bit nodes exist, which bounds decoding error probability away from 0.

46 We Propose Using Coupled HA Codes for Rateless Coding x d r Shortened {}}{ 0 {}}{ L 1 {}}{ Shortened {}}{ x d l e l avg(x 1) x d g Do the codes achieve capacity of BEC or achieve zero-overhead? If so how small the degree d l, d r, d g can be?

47 Density Evolution Predicts Error Probability i: section and l: BP iteration round p (l) i : erasure probability of BP messages sent from bit nodes to check nodes s (l) i : erasure probability of BP messages sent from bit nodes to channel nodes p (0) i = s (0) i = ( p (l+1) 1 i = w ( 1 Λ w ( s (l+1) 1 i = w ( 1 Λ w w 1 j=0 w 1 j=0 w 1 j=0 w 1 j=0 { 0 i / [0, L 1] 1 i [0, L 1] ( 1 (1 1 w w 1 k=0 ( 1 (1 ϵ)(1 1 w ( 1 (1 1 w w 1 k=0 ( 1 (1 ϵ)(1 1 w Λ(x) = exp[ l avg (1 x)], p (l) i+j k )d r 1 )) d l 1 w 1 k=0 s (l) i+j k )d g 1 )), p (l) i+j k )dr 1)) d l w 1 s (l) i+j k )d g 1 ))

48 We are interested in minimum overhead α such that decoding erasure probability goes to 0. From density evolution, overhead threshold αl is calculated as follows α L := inf { α > 0 P ( ) b (L) = 0 }, P (l) b l avg = := 1 L L 1 p (l) i, i=0 d g 1 ϵ LR pre (L)(1 + α) L + w 1 From the rateloss effect, we need large L to have α L = 0 Do the proposed codes achieve zero overhead threshold lim L α L = 0 If so how small the degree d l, d r, d g can be?

49 Observed proposed codes achieve zero overhead threshold with some parameters d l, d r, d g. How small can they be? We numerically observed that the proposed codes achieve zero overhead threshold lim L α L = 0 if d l 3, d g 3 and d r 4. When d l = 2, we numerically observed that the proposed codes achieve zero overhead threshold with some parameter tuples (d l = 2, d r, d g ). Overhead Threshold α L 0.5 d l = 2, d r = 4, d g = 2 d l = 2, d r = 3, d g = 2 d l = 2, d r = 3, d g = 3 d l = 2, d r = 4, d g = L

50 Necessary Condition of Capacity-Achieving Coupled Precoded Regular Rateless Codes Theorem: For capacity-achieving (d l = 2, d r, d g, L, w) precoded rateless codes have to satisfy d g d r ln(d r 1). d r 2 This implies d r > 2 or d g > 2 is necessary. Outline of Proof: Taylor expansion around 0 gives p (l+1) = P L p (l) + o(p (l) ), where P L is an L L band matrix as follows { w i j (P L ) i,j = w (d 2 r 1)λ(ϵ) ( i j w) 0 ( i j > w) Evaluate the spectrum radius ρ(p L ) of P L gives a necessary condition. x T P L x 1 > ρ(p L ) = max x R L :x 0 x T x 1T P L 1 1 T 1 L = (dr 1)e (l avg) L (1 ϵ),

51 Conclusion Achievement Empirically observed that precoded rateless codes achieve the capacity over the BEC. Derive a necessary condition of capacity-achieving codes. Ongoing work Prove proposed codes achieve capacity of BEC via potential function. (completed) An extension to BMS channels and a proof for universal capacity-achievability.

Message Passing Algorithm with MAP Decoding on Zigzag Cycles for Non-binary LDPC Codes

Message Passing Algorithm with MAP Decoding on Zigzag Cycles for Non-binary LDPC Codes Message Passing Algorithm with MAP Decoding on Zigzag Cycles for Non-binary LDPC Codes Takayuki Nozaki 1, Kenta Kasai 2, Kohichi Sakaniwa 2 1 Kanagawa University 2 Tokyo Institute of Technology July 12th,

More information

Lower Bounds on the Graphical Complexity of Finite-Length LDPC Codes

Lower Bounds on the Graphical Complexity of Finite-Length LDPC Codes Lower Bounds on the Graphical Complexity of Finite-Length LDPC Codes Igal Sason Department of Electrical Engineering Technion - Israel Institute of Technology Haifa 32000, Israel 2009 IEEE International

More information

ECEN 655: Advanced Channel Coding

ECEN 655: Advanced Channel Coding ECEN 655: Advanced Channel Coding Course Introduction Henry D. Pfister Department of Electrical and Computer Engineering Texas A&M University ECEN 655: Advanced Channel Coding 1 / 19 Outline 1 History

More information

Spatially-Coupled MacKay-Neal Codes Universally Achieve the Symmetric Information Rate of Arbitrary Generalized Erasure Channels with Memory

Spatially-Coupled MacKay-Neal Codes Universally Achieve the Symmetric Information Rate of Arbitrary Generalized Erasure Channels with Memory Spatially-Coupled MacKay-Neal Codes Universally Achieve the Symmetric Information Rate of Arbitrary Generalized Erasure Channels with Memory arxiv:1516736v1 [csit] 27 Jan 215 Abstract This paper investigates

More information

LDPC Codes. Slides originally from I. Land p.1

LDPC Codes. Slides originally from I. Land p.1 Slides originally from I. Land p.1 LDPC Codes Definition of LDPC Codes Factor Graphs to use in decoding Decoding for binary erasure channels EXIT charts Soft-Output Decoding Turbo principle applied to

More information

Threshold Saturation on Channels with Memory via Spatial Coupling

Threshold Saturation on Channels with Memory via Spatial Coupling Threshold Saturation on Channels with Memory via Spatial Coupling Shrinivas Kudekar and Kenta Kasai New Mexico Consortium and Center for Non-linear Studies, Los Alamos National Laboratory, NM, USA Email:

More information

Performance Analysis and Code Optimization of Low Density Parity-Check Codes on Rayleigh Fading Channels

Performance Analysis and Code Optimization of Low Density Parity-Check Codes on Rayleigh Fading Channels Performance Analysis and Code Optimization of Low Density Parity-Check Codes on Rayleigh Fading Channels Jilei Hou, Paul H. Siegel and Laurence B. Milstein Department of Electrical and Computer Engineering

More information

Spatially Coupled LDPC Codes Constructed from Protographs

Spatially Coupled LDPC Codes Constructed from Protographs IEEE TRANSACTIONS ON INFORMATION THEORY (SUBMITTED PAPER) 1 Spatially Coupled LDPC Codes Constructed from Protographs David G. M. Mitchell, Member, IEEE, Michael Lentmaier, Senior Member, IEEE, and Daniel

More information

Lecture 4 : Introduction to Low-density Parity-check Codes

Lecture 4 : Introduction to Low-density Parity-check Codes Lecture 4 : Introduction to Low-density Parity-check Codes LDPC codes are a class of linear block codes with implementable decoders, which provide near-capacity performance. History: 1. LDPC codes were

More information

Time-invariant LDPC convolutional codes

Time-invariant LDPC convolutional codes Time-invariant LDPC convolutional codes Dimitris Achlioptas, Hamed Hassani, Wei Liu, and Rüdiger Urbanke Department of Computer Science, UC Santa Cruz, USA Email: achlioptas@csucscedu Department of Computer

More information

EE229B - Final Project. Capacity-Approaching Low-Density Parity-Check Codes

EE229B - Final Project. Capacity-Approaching Low-Density Parity-Check Codes EE229B - Final Project Capacity-Approaching Low-Density Parity-Check Codes Pierre Garrigues EECS department, UC Berkeley garrigue@eecs.berkeley.edu May 13, 2005 Abstract The class of low-density parity-check

More information

Capacity-Achieving Ensembles for the Binary Erasure Channel With Bounded Complexity

Capacity-Achieving Ensembles for the Binary Erasure Channel With Bounded Complexity Capacity-Achieving Ensembles for the Binary Erasure Channel With Bounded Complexity Henry D. Pfister, Member, Igal Sason, Member, and Rüdiger Urbanke Abstract We present two sequences of ensembles of non-systematic

More information

Quasi-Cyclic Asymptotically Regular LDPC Codes

Quasi-Cyclic Asymptotically Regular LDPC Codes 2010 IEEE Information Theory Workshop - ITW 2010 Dublin Quasi-Cyclic Asymptotically Regular LDPC Codes David G. M. Mitchell, Roxana Smarandache, Michael Lentmaier, and Daniel J. Costello, Jr. Dept. of

More information

On Achievable Rates and Complexity of LDPC Codes over Parallel Channels: Bounds and Applications

On Achievable Rates and Complexity of LDPC Codes over Parallel Channels: Bounds and Applications On Achievable Rates and Complexity of LDPC Codes over Parallel Channels: Bounds and Applications Igal Sason, Member and Gil Wiechman, Graduate Student Member Abstract A variety of communication scenarios

More information

Accumulate-Repeat-Accumulate Codes: Capacity-Achieving Ensembles of Systematic Codes for the Erasure Channel with Bounded Complexity

Accumulate-Repeat-Accumulate Codes: Capacity-Achieving Ensembles of Systematic Codes for the Erasure Channel with Bounded Complexity Accumulate-Repeat-Accumulate Codes: Capacity-Achieving Ensembles of Systematic Codes for the Erasure Channel with Bounded Complexity Henry D. Pfister, Member, Igal Sason, Member Abstract The paper introduces

More information

Staircase Codes. for High-Speed Optical Communications

Staircase Codes. for High-Speed Optical Communications Staircase Codes for High-Speed Optical Communications Frank R. Kschischang Dept. of Electrical & Computer Engineering University of Toronto (Joint work with Lei Zhang, Benjamin Smith, Arash Farhood, Andrew

More information

Belief-Propagation Decoding of LDPC Codes

Belief-Propagation Decoding of LDPC Codes LDPC Codes: Motivation Belief-Propagation Decoding of LDPC Codes Amir Bennatan, Princeton University Revolution in coding theory Reliable transmission, rates approaching capacity. BIAWGN, Rate =.5, Threshold.45

More information

On Generalized EXIT Charts of LDPC Code Ensembles over Binary-Input Output-Symmetric Memoryless Channels

On Generalized EXIT Charts of LDPC Code Ensembles over Binary-Input Output-Symmetric Memoryless Channels 2012 IEEE International Symposium on Information Theory Proceedings On Generalied EXIT Charts of LDPC Code Ensembles over Binary-Input Output-Symmetric Memoryless Channels H Mamani 1, H Saeedi 1, A Eslami

More information

SPARSE graph codes have been of great interest in the

SPARSE graph codes have been of great interest in the Windowed Decoding of Spatially Coupled Codes Aravind R. Iyengar, Paul H. Siegel, Fellow, IEEE, Rüdiger L. Urbanke and Jack K. Wolf, Life Fellow, IEEE arxiv:06.0075v3 [cs.it] 30 Nov 0 Abstract Spatially

More information

Symmetric Product Codes

Symmetric Product Codes Symmetric Product Codes Henry D. Pfister 1, Santosh Emmadi 2, and Krishna Narayanan 2 1 Department of Electrical and Computer Engineering Duke University 2 Department of Electrical and Computer Engineering

More information

Convergence analysis for a class of LDPC convolutional codes on the erasure channel

Convergence analysis for a class of LDPC convolutional codes on the erasure channel Convergence analysis for a class of LDPC convolutional codes on the erasure channel Sridharan, Arvind; Lentmaier, Michael; Costello Jr., Daniel J.; Zigangirov, Kamil Published in: [Host publication title

More information

Recent Results on Capacity-Achieving Codes for the Erasure Channel with Bounded Complexity

Recent Results on Capacity-Achieving Codes for the Erasure Channel with Bounded Complexity 26 IEEE 24th Convention of Electrical and Electronics Engineers in Israel Recent Results on Capacity-Achieving Codes for the Erasure Channel with Bounded Complexity Igal Sason Technion Israel Institute

More information

Research Collection. The fractality of polar codes. Conference Paper. ETH Library. Author(s): Geiger, Bernhard C. Publication Date: 2016

Research Collection. The fractality of polar codes. Conference Paper. ETH Library. Author(s): Geiger, Bernhard C. Publication Date: 2016 Research Collection Conference Paper The fractality of polar codes Authors: Geiger, ernhard C. Publication Date: 016 Permanent Link: https://doi.org/10.399/ethz-a-01064591 Rights / License: In Copyright

More information

Spatially-coupled Split-component Codes with Iterative Algebraic Decoding

Spatially-coupled Split-component Codes with Iterative Algebraic Decoding Spatially-coupled Split-component Codes with Iterative Algebraic Decoding Lei M. Zhang, Student Member, IEEE Dmitri Truhachev, Member, IEEE and Frank R. Kschischang, Fellow, IEEE 1 arxiv:1512.01132v1 [cs.it]

More information

Codes designed via algebraic lifts of graphs

Codes designed via algebraic lifts of graphs p./40 Codes designed via algebraic lifts of graphs Clemson Mini-Conference on Discrete Mathematics Oct. 3, 2008. Christine A. Kelley Department of Mathematics University of Nebraska-Lincoln email: ckelley2@math.unl.edu

More information

Graph-based codes for flash memory

Graph-based codes for flash memory 1/28 Graph-based codes for flash memory Discrete Mathematics Seminar September 3, 2013 Katie Haymaker Joint work with Professor Christine Kelley University of Nebraska-Lincoln 2/28 Outline 1 Background

More information

LDPC Codes. Intracom Telecom, Peania

LDPC Codes. Intracom Telecom, Peania LDPC Codes Alexios Balatsoukas-Stimming and Athanasios P. Liavas Technical University of Crete Dept. of Electronic and Computer Engineering Telecommunications Laboratory December 16, 2011 Intracom Telecom,

More information

THIS paper provides a general technique for constructing

THIS paper provides a general technique for constructing Protograph-Based Raptor-Like LDPC Codes for the Binary Erasure Channel Kasra Vakilinia Department of Electrical Engineering University of California, Los Angeles Los Angeles, California 90024 Email: vakiliniak@ucla.edu

More information

Low-Density Parity-Check Codes

Low-Density Parity-Check Codes Department of Computer Sciences Applied Algorithms Lab. July 24, 2011 Outline 1 Introduction 2 Algorithms for LDPC 3 Properties 4 Iterative Learning in Crowds 5 Algorithm 6 Results 7 Conclusion PART I

More information

LDPC Code Ensembles that Universally Achieve Capacity under BP Decoding: A Simple Derivation

LDPC Code Ensembles that Universally Achieve Capacity under BP Decoding: A Simple Derivation LDPC Code Ensembles that Universally Achieve Capacity under BP Decoding: A Simple Derivation Anatoly Khina EE Systems Dept., TAU Tel Aviv, Israel Email: anatolyk@eng.tau.ac.il Yair Yona Dept. of EE, UCLA

More information

An Introduction to Low Density Parity Check (LDPC) Codes

An Introduction to Low Density Parity Check (LDPC) Codes An Introduction to Low Density Parity Check (LDPC) Codes Jian Sun jian@csee.wvu.edu Wireless Communication Research Laboratory Lane Dept. of Comp. Sci. and Elec. Engr. West Virginia University June 3,

More information

Introduction to Low-Density Parity Check Codes. Brian Kurkoski

Introduction to Low-Density Parity Check Codes. Brian Kurkoski Introduction to Low-Density Parity Check Codes Brian Kurkoski kurkoski@ice.uec.ac.jp Outline: Low Density Parity Check Codes Review block codes History Low Density Parity Check Codes Gallager s LDPC code

More information

Codes on Graphs. Telecommunications Laboratory. Alex Balatsoukas-Stimming. Technical University of Crete. November 27th, 2008

Codes on Graphs. Telecommunications Laboratory. Alex Balatsoukas-Stimming. Technical University of Crete. November 27th, 2008 Codes on Graphs Telecommunications Laboratory Alex Balatsoukas-Stimming Technical University of Crete November 27th, 2008 Telecommunications Laboratory (TUC) Codes on Graphs November 27th, 2008 1 / 31

More information

Upper Bounding the Performance of Arbitrary Finite LDPC Codes on Binary Erasure Channels

Upper Bounding the Performance of Arbitrary Finite LDPC Codes on Binary Erasure Channels Upper Bounding the Performance of Arbitrary Finite LDPC Codes on Binary Erasure Channels An efficient exhaustion algorithm for error-prone patterns C.-C. Wang, S.R. Kulkarni, and H.V. Poor School of Electrical

More information

Codes on graphs and iterative decoding

Codes on graphs and iterative decoding Codes on graphs and iterative decoding Bane Vasić Error Correction Coding Laboratory University of Arizona Prelude Information transmission 0 0 0 0 0 0 Channel Information transmission signal 0 0 threshold

More information

Fountain Codes. Amin Shokrollahi EPFL

Fountain Codes. Amin Shokrollahi EPFL Fountain Codes Amin Shokrollahi EPFL Content Fountain Codes LT-Codes Raptor Codes Extensions Transmission Problem Transmit information from one sender to multiple receivers where the channel between sender

More information

Enhancing Binary Images of Non-Binary LDPC Codes

Enhancing Binary Images of Non-Binary LDPC Codes Enhancing Binary Images of Non-Binary LDPC Codes Aman Bhatia, Aravind R Iyengar, and Paul H Siegel University of California, San Diego, La Jolla, CA 92093 0401, USA Email: {a1bhatia, aravind, psiegel}@ucsdedu

More information

On Bit Error Rate Performance of Polar Codes in Finite Regime

On Bit Error Rate Performance of Polar Codes in Finite Regime On Bit Error Rate Performance of Polar Codes in Finite Regime A. Eslami and H. Pishro-Nik Abstract Polar codes have been recently proposed as the first low complexity class of codes that can provably achieve

More information

Joint Iterative Decoding of LDPC Codes and Channels with Memory

Joint Iterative Decoding of LDPC Codes and Channels with Memory Joint Iterative Decoding of LDPC Codes and Channels with Memory Henry D. Pfister and Paul H. Siegel University of California, San Diego 3 rd Int l Symposium on Turbo Codes September 1, 2003 Outline Channels

More information

Iterative Encoding of Low-Density Parity-Check Codes

Iterative Encoding of Low-Density Parity-Check Codes Iterative Encoding of Low-Density Parity-Check Codes David Haley, Alex Grant and John Buetefuer Institute for Telecommunications Research University of South Australia Mawson Lakes Blvd Mawson Lakes SA

More information

BOUNDS ON THE MAP THRESHOLD OF ITERATIVE DECODING SYSTEMS WITH ERASURE NOISE. A Thesis CHIA-WEN WANG

BOUNDS ON THE MAP THRESHOLD OF ITERATIVE DECODING SYSTEMS WITH ERASURE NOISE. A Thesis CHIA-WEN WANG BOUNDS ON THE MAP THRESHOLD OF ITERATIVE DECODING SYSTEMS WITH ERASURE NOISE A Thesis by CHIA-WEN WANG Submitted to the Office of Graduate Studies of Texas A&M University in partial fulfillment of the

More information

Asynchronous Decoding of LDPC Codes over BEC

Asynchronous Decoding of LDPC Codes over BEC Decoding of LDPC Codes over BEC Saeid Haghighatshoar, Amin Karbasi, Amir Hesam Salavati Department of Telecommunication Systems, Technische Universität Berlin, Germany, School of Engineering and Applied

More information

Ma/CS 6b Class 25: Error Correcting Codes 2

Ma/CS 6b Class 25: Error Correcting Codes 2 Ma/CS 6b Class 25: Error Correcting Codes 2 By Adam Sheffer Recall: Codes V n the set of binary sequences of length n. For example, V 3 = 000,001,010,011,100,101,110,111. Codes of length n are subsets

More information

arxiv:cs/ v1 [cs.it] 16 Aug 2005

arxiv:cs/ v1 [cs.it] 16 Aug 2005 On Achievable Rates and Complexity of LDPC Codes for Parallel Channels with Application to Puncturing Igal Sason Gil Wiechman arxiv:cs/587v [cs.it] 6 Aug 5 Technion Israel Institute of Technology Haifa

More information

Turbo Codes are Low Density Parity Check Codes

Turbo Codes are Low Density Parity Check Codes Turbo Codes are Low Density Parity Check Codes David J. C. MacKay July 5, 00 Draft 0., not for distribution! (First draft written July 5, 998) Abstract Turbo codes and Gallager codes (also known as low

More information

Construction of LDPC codes

Construction of LDPC codes Construction of LDPC codes Telecommunications Laboratory Alex Balatsoukas-Stimming Technical University of Crete July 1, 2009 Telecommunications Laboratory (TUC) Construction of LDPC codes July 1, 2009

More information

CHAPTER 3 LOW DENSITY PARITY CHECK CODES

CHAPTER 3 LOW DENSITY PARITY CHECK CODES 62 CHAPTER 3 LOW DENSITY PARITY CHECK CODES 3. INTRODUCTION LDPC codes were first presented by Gallager in 962 [] and in 996, MacKay and Neal re-discovered LDPC codes.they proved that these codes approach

More information

Accumulate-Repeat-Accumulate Codes: Capacity-Achieving Ensembles of Systematic Codes for the Erasure Channel with Bounded Complexity

Accumulate-Repeat-Accumulate Codes: Capacity-Achieving Ensembles of Systematic Codes for the Erasure Channel with Bounded Complexity Accumulate-Repeat-Accumulate Codes: Capacity-Achieving Ensembles of Systematic Codes for the Erasure Channel with Bounded Complexity Henry D. Pfister, Member, Igal Sason, Member Abstract This paper introduces

More information

Design and Analysis of Graph-based Codes Using Algebraic Lifts and Decoding Networks

Design and Analysis of Graph-based Codes Using Algebraic Lifts and Decoding Networks University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln Dissertations, Theses, and Student Research Papers in Mathematics Mathematics, Department of 3-2018 Design and Analysis

More information

Raptor Codes: From a Math Idea to LTE embms. BIRS, October 2015

Raptor Codes: From a Math Idea to LTE embms. BIRS, October 2015 Raptor Codes: From a Math Idea to LTE embms BIRS, October 2015 The plan is to... 1 introduce LT codes and Raptor codes 2 provide insights into their design 3 address some common misconceptions 2 / 31 The

More information

Binary Convolutional Codes

Binary Convolutional Codes Binary Convolutional Codes A convolutional code has memory over a short block length. This memory results in encoded output symbols that depend not only on the present input, but also on past inputs. An

More information

An Introduction to Low-Density Parity-Check Codes

An Introduction to Low-Density Parity-Check Codes An Introduction to Low-Density Parity-Check Codes Paul H. Siegel Electrical and Computer Engineering University of California, San Diego 5/ 3/ 7 Copyright 27 by Paul H. Siegel Outline Shannon s Channel

More information

Modern Coding Theory. Daniel J. Costello, Jr School of Information Theory Northwestern University August 10, 2009

Modern Coding Theory. Daniel J. Costello, Jr School of Information Theory Northwestern University August 10, 2009 Modern Coding Theory Daniel J. Costello, Jr. Coding Research Group Department of Electrical Engineering University of Notre Dame Notre Dame, IN 46556 2009 School of Information Theory Northwestern University

More information

Capacity of a channel Shannon s second theorem. Information Theory 1/33

Capacity of a channel Shannon s second theorem. Information Theory 1/33 Capacity of a channel Shannon s second theorem Information Theory 1/33 Outline 1. Memoryless channels, examples ; 2. Capacity ; 3. Symmetric channels ; 4. Channel Coding ; 5. Shannon s second theorem,

More information

On the Typicality of the Linear Code Among the LDPC Coset Code Ensemble

On the Typicality of the Linear Code Among the LDPC Coset Code Ensemble 5 Conference on Information Sciences and Systems The Johns Hopkins University March 16 18 5 On the Typicality of the Linear Code Among the LDPC Coset Code Ensemble C.-C. Wang S.R. Kulkarni and H.V. Poor

More information

Belief propagation decoding of quantum channels by passing quantum messages

Belief propagation decoding of quantum channels by passing quantum messages Belief propagation decoding of quantum channels by passing quantum messages arxiv:67.4833 QIP 27 Joseph M. Renes lempelziv@flickr To do research in quantum information theory, pick a favorite text on classical

More information

Fountain Uncorrectable Sets and Finite-Length Analysis

Fountain Uncorrectable Sets and Finite-Length Analysis Fountain Uncorrectable Sets and Finite-Length Analysis Wen Ji 1, Bo-Wei Chen 2, and Yiqiang Chen 1 1 Beijing Key Laboratory of Mobile Computing and Pervasive Device Institute of Computing Technology, Chinese

More information

How to Achieve the Capacity of Asymmetric Channels

How to Achieve the Capacity of Asymmetric Channels How to Achieve the Capacity of Asymmetric Channels Marco Mondelli, S. Hamed Hassani, and Rüdiger Urbanke Abstract arxiv:406.7373v5 [cs.it] 3 Jan 208 We survey coding techniques that enable reliable transmission

More information

GLDPC-Staircase AL-FEC codes: A Fundamental study and New results

GLDPC-Staircase AL-FEC codes: A Fundamental study and New results GLDPC-Staircase AL-FEC codes: A Fundamental study and New results Ferdaouss Mattoussi, Vincent Roca, Bessam Sayadi To cite this version: Ferdaouss Mattoussi, Vincent Roca, Bessam Sayadi. GLDPC-Staircase

More information

From Stopping sets to Trapping sets

From Stopping sets to Trapping sets From Stopping sets to Trapping sets The Exhaustive Search Algorithm & The Suppressing Effect Chih-Chun Wang School of Electrical & Computer Engineering Purdue University Wang p. 1/21 Content Good exhaustive

More information

Capacity-approaching codes

Capacity-approaching codes Chapter 13 Capacity-approaching codes We have previously discussed codes on graphs and the sum-product decoding algorithm in general terms. In this chapter we will give a brief overview of some particular

More information

Sparse Superposition Codes for the Gaussian Channel

Sparse Superposition Codes for the Gaussian Channel Sparse Superposition Codes for the Gaussian Channel Florent Krzakala (LPS, Ecole Normale Supérieure, France) J. Barbier (ENS) arxiv:1403.8024 presented at ISIT 14 Long version in preparation Communication

More information

Practical Polar Code Construction Using Generalised Generator Matrices

Practical Polar Code Construction Using Generalised Generator Matrices Practical Polar Code Construction Using Generalised Generator Matrices Berksan Serbetci and Ali E. Pusane Department of Electrical and Electronics Engineering Bogazici University Istanbul, Turkey E-mail:

More information

Quasi-cyclic Low Density Parity Check codes with high girth

Quasi-cyclic Low Density Parity Check codes with high girth Quasi-cyclic Low Density Parity Check codes with high girth, a work with Marta Rossi, Richard Bresnan, Massimilliano Sala Summer Doctoral School 2009 Groebner bases, Geometric codes and Order Domains Dept

More information

An Efficient Maximum Likelihood Decoding of LDPC Codes Over the Binary Erasure Channel

An Efficient Maximum Likelihood Decoding of LDPC Codes Over the Binary Erasure Channel IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 5, NO. 11, NOVEMBER 24 1 An Efficient Maximum Likelihood Decoding of LDPC Codes Over the Binary Erasure Channel David Burshtein and Gadi Miller School of Electrical

More information

Girth Analysis of Polynomial-Based Time-Invariant LDPC Convolutional Codes

Girth Analysis of Polynomial-Based Time-Invariant LDPC Convolutional Codes IWSSIP 212, 11-13 April 212, Vienna, Austria ISBN 978-3-2-2328-4 Girth Analysis of Polynomial-Based Time-Invariant LDPC Convolutional Codes Hua Zhou and Norbert Goertz Institute of Telecommunications Vienna

More information

Polar Codes: Graph Representation and Duality

Polar Codes: Graph Representation and Duality Polar Codes: Graph Representation and Duality arxiv:1312.0372v1 [cs.it] 2 Dec 2013 M. Fossorier ETIS ENSEA/UCP/CNRS UMR-8051 6, avenue du Ponceau, 95014, Cergy Pontoise, France Email: mfossorier@ieee.org

More information

Low-complexity error correction in LDPC codes with constituent RS codes 1

Low-complexity error correction in LDPC codes with constituent RS codes 1 Eleventh International Workshop on Algebraic and Combinatorial Coding Theory June 16-22, 2008, Pamporovo, Bulgaria pp. 348-353 Low-complexity error correction in LDPC codes with constituent RS codes 1

More information

5. Density evolution. Density evolution 5-1

5. Density evolution. Density evolution 5-1 5. Density evolution Density evolution 5-1 Probabilistic analysis of message passing algorithms variable nodes factor nodes x1 a x i x2 a(x i ; x j ; x k ) x3 b x4 consider factor graph model G = (V ;

More information

Iterative Quantization. Using Codes On Graphs

Iterative Quantization. Using Codes On Graphs Iterative Quantization Using Codes On Graphs Emin Martinian and Jonathan S. Yedidia 2 Massachusetts Institute of Technology 2 Mitsubishi Electric Research Labs Lossy Data Compression: Encoding: Map source

More information

arxiv: v1 [eess.sp] 22 Oct 2017

arxiv: v1 [eess.sp] 22 Oct 2017 A Binary Wyner-Ziv Code Design Based on Compound LDGM-LDPC Structures Mahdi Nangir 1, Mahmoud Ahmadian-Attari 1, and Reza Asvadi 2,* arxiv:1710.07985v1 [eess.sp] 22 Oct 2017 1 Faculty of Electrical Engineering,

More information

Low Density Parity Check (LDPC) Codes and the Need for Stronger ECC. August 2011 Ravi Motwani, Zion Kwok, Scott Nelson

Low Density Parity Check (LDPC) Codes and the Need for Stronger ECC. August 2011 Ravi Motwani, Zion Kwok, Scott Nelson Low Density Parity Check (LDPC) Codes and the Need for Stronger ECC August 2011 Ravi Motwani, Zion Kwok, Scott Nelson Agenda NAND ECC History Soft Information What is soft information How do we obtain

More information

Part III Advanced Coding Techniques

Part III Advanced Coding Techniques Part III Advanced Coding Techniques José Vieira SPL Signal Processing Laboratory Departamento de Electrónica, Telecomunicações e Informática / IEETA Universidade de Aveiro, Portugal 2010 José Vieira (IEETA,

More information

Lecture 2 Linear Codes

Lecture 2 Linear Codes Lecture 2 Linear Codes 2.1. Linear Codes From now on we want to identify the alphabet Σ with a finite field F q. For general codes, introduced in the last section, the description is hard. For a code of

More information

Lecture 8: Shannon s Noise Models

Lecture 8: Shannon s Noise Models Error Correcting Codes: Combinatorics, Algorithms and Applications (Fall 2007) Lecture 8: Shannon s Noise Models September 14, 2007 Lecturer: Atri Rudra Scribe: Sandipan Kundu& Atri Rudra Till now we have

More information

Polar Codes are Optimal for Lossy Source Coding

Polar Codes are Optimal for Lossy Source Coding Polar Codes are Optimal for Lossy Source Coding Satish Babu Korada and Rüdiger Urbanke EPFL, Switzerland, Email: satish.korada,ruediger.urbanke}@epfl.ch Abstract We consider lossy source compression of

More information

Lecture 3: Channel Capacity

Lecture 3: Channel Capacity Lecture 3: Channel Capacity 1 Definitions Channel capacity is a measure of maximum information per channel usage one can get through a channel. This one of the fundamental concepts in information theory.

More information

Constructions of Nonbinary Quasi-Cyclic LDPC Codes: A Finite Field Approach

Constructions of Nonbinary Quasi-Cyclic LDPC Codes: A Finite Field Approach Constructions of Nonbinary Quasi-Cyclic LDPC Codes: A Finite Field Approach Shu Lin, Shumei Song, Lan Lan, Lingqi Zeng and Ying Y Tai Department of Electrical & Computer Engineering University of California,

More information

Growth Rate of Spatially Coupled LDPC codes

Growth Rate of Spatially Coupled LDPC codes Growth Rate of Spatially Coupled LDPC codes Workshop on Spatially Coupled Codes and Related Topics at Tokyo Institute of Technology 2011/2/19 Contents 1. Factor graph, Bethe approximation and belief propagation

More information

Erasure Codes with a Banded Structure for Hybrid Iterative-ML Decoding

Erasure Codes with a Banded Structure for Hybrid Iterative-ML Decoding Erasure Codes with a Banded Structure for Hybrid Iterative-ML Decoding Alexandre Soro 1, Mathieu Cunche 2, Jérôme Lacan 1 and Vincent Roca 2 1 Univ of Toulouse, ISAE/DMIA, 10 avenue Edouard Belin, BP 54032-31055

More information

Codes on graphs and iterative decoding

Codes on graphs and iterative decoding Codes on graphs and iterative decoding Bane Vasić Error Correction Coding Laboratory University of Arizona Funded by: National Science Foundation (NSF) Seagate Technology Defense Advanced Research Projects

More information

Distance Properties of Short LDPC Codes and Their Impact on the BP, ML and Near-ML Decoding Performance

Distance Properties of Short LDPC Codes and Their Impact on the BP, ML and Near-ML Decoding Performance Distance Properties of Short LDPC Codes and Their Impact on the BP, ML and Near-ML Decoding Performance Irina E. Bocharova 1,2, Boris D. Kudryashov 1, Vitaly Skachek 2, Yauhen Yakimenka 2 1 St. Petersburg

More information

Globally Coupled LDPC Codes

Globally Coupled LDPC Codes Globally Coupled LDPC Codes Juane Li 1, Shu Lin 1, Khaled Abdel-Ghaffar 1, William E Ryan 2, and Daniel J Costello, Jr 3 1 University of California, Davis, CA 95616 2 Zeta Associates, Fairfax, VA 22030

More information

Low-density parity-check codes

Low-density parity-check codes Low-density parity-check codes From principles to practice Dr. Steve Weller steven.weller@newcastle.edu.au School of Electrical Engineering and Computer Science The University of Newcastle, Callaghan,

More information

ML and Near-ML Decoding of LDPC Codes Over the BEC: Bounds and Decoding Algorithms

ML and Near-ML Decoding of LDPC Codes Over the BEC: Bounds and Decoding Algorithms 1 ML and Near-ML Decoding of LDPC Codes Over the BEC: Bounds and Decoding Algorithms Irina E. Bocharova, Senior Member, IEEE, Boris D. Kudryashov, Senior Member, IEEE, Vitaly Skachek, Member, IEEE, Eirik

More information

An algorithm to improve the error rate performance of Accumulate-Repeat-Accumulate codes Tae-Ui Kim

An algorithm to improve the error rate performance of Accumulate-Repeat-Accumulate codes Tae-Ui Kim An algorithm to improve the error rate performance of Accumulate-Repeat-Accumulate codes Tae-Ui Kim The Graduate School Yonsei University Department of Electrical and Electronic Engineering An algorithm

More information

Variable Rate Channel Capacity. Jie Ren 2013/4/26

Variable Rate Channel Capacity. Jie Ren 2013/4/26 Variable Rate Channel Capacity Jie Ren 2013/4/26 Reference This is a introduc?on of Sergio Verdu and Shlomo Shamai s paper. Sergio Verdu and Shlomo Shamai, Variable- Rate Channel Capacity, IEEE Transac?ons

More information

Iterative Decoding for Wireless Networks

Iterative Decoding for Wireless Networks Iterative Decoding for Wireless Networks Thesis by Ravi Palanki In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California

More information

An Efficient Algorithm for Finding Dominant Trapping Sets of LDPC Codes

An Efficient Algorithm for Finding Dominant Trapping Sets of LDPC Codes An Efficient Algorithm for Finding Dominant Trapping Sets of LDPC Codes Mehdi Karimi, Student Member, IEEE and Amir H. Banihashemi, Senior Member, IEEE Abstract arxiv:1108.4478v2 [cs.it] 13 Apr 2012 This

More information

Lecture 9 Polar Coding

Lecture 9 Polar Coding Lecture 9 Polar Coding I-Hsiang ang Department of Electrical Engineering National Taiwan University ihwang@ntu.edu.tw December 29, 2015 1 / 25 I-Hsiang ang IT Lecture 9 In Pursuit of Shannon s Limit Since

More information

Unified Scaling of Polar Codes: Error Exponent, Scaling Exponent, Moderate Deviations, and Error Floors

Unified Scaling of Polar Codes: Error Exponent, Scaling Exponent, Moderate Deviations, and Error Floors Unified Scaling of Polar Codes: Error Exponent, Scaling Exponent, Moderate Deviations, and Error Floors Marco Mondelli, S. Hamed Hassani, and Rüdiger Urbanke Abstract Consider transmission of a polar code

More information

An Introduction to Algorithmic Coding Theory

An Introduction to Algorithmic Coding Theory An Introduction to Algorithmic Coding Theory M. Amin Shokrollahi Bell Laboratories Part : Codes - A puzzle What do the following problems have in common? 2 Problem : Information Transmission MESSAGE G

More information

Low-density parity-check (LDPC) codes

Low-density parity-check (LDPC) codes Low-density parity-check (LDPC) codes Performance similar to turbo codes Do not require long interleaver to achieve good performance Better block error performance Error floor occurs at lower BER Decoding

More information

On Concentration of Martingales and Applications in Information Theory, Communication & Coding

On Concentration of Martingales and Applications in Information Theory, Communication & Coding On Concentration of Martingales and Applications in Information Theory, Communication & Coding Igal Sason Department of Electrical Engineering Technion - Israel Institute of Technology Haifa 32000, Israel

More information

The PPM Poisson Channel: Finite-Length Bounds and Code Design

The PPM Poisson Channel: Finite-Length Bounds and Code Design August 21, 2014 The PPM Poisson Channel: Finite-Length Bounds and Code Design Flavio Zabini DEI - University of Bologna and Institute for Communications and Navigation German Aerospace Center (DLR) Balazs

More information

Optimizing Flash based Storage Systems

Optimizing Flash based Storage Systems Optimizing Flash based Storage Systems 1 Lifetime Reliability Latency Throughput 2 Projects Reliability/Latency/Throughput: Error Correction Code (ECC) Parallelization and Incremental Redundancy Lifetime:

More information

Introducing Low-Density Parity-Check Codes

Introducing Low-Density Parity-Check Codes Introducing Low-Density Parity-Check Codes Sarah J. Johnson School of Electrical Engineering and Computer Science The University of Newcastle Australia email: sarah.johnson@newcastle.edu.au Topic 1: Low-Density

More information

Coding Techniques for Data Storage Systems

Coding Techniques for Data Storage Systems Coding Techniques for Data Storage Systems Thomas Mittelholzer IBM Zurich Research Laboratory /8 Göttingen Agenda. Channel Coding and Practical Coding Constraints. Linear Codes 3. Weight Enumerators and

More information

Optimal Rate and Maximum Erasure Probability LDPC Codes in Binary Erasure Channel

Optimal Rate and Maximum Erasure Probability LDPC Codes in Binary Erasure Channel Optimal Rate and Maximum Erasure Probability LDPC Codes in Binary Erasure Channel H. Tavakoli Electrical Engineering Department K.N. Toosi University of Technology, Tehran, Iran tavakoli@ee.kntu.ac.ir

More information

Compressed Sensing and Linear Codes over Real Numbers

Compressed Sensing and Linear Codes over Real Numbers Compressed Sensing and Linear Codes over Real Numbers Henry D. Pfister (joint with Fan Zhang) Texas A&M University College Station Information Theory and Applications Workshop UC San Diego January 31st,

More information