1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118
//! # Progressive Edge Growth (PEG) LDPC construction
//!
//! This implements the algorithm described in *Xiao-Yu Hu, E. Eleftheriou and
//! D. M. Arnold, "Regular and irregular progressive edge-growth tanner graphs,"
//! in IEEE Transactions on Information Theory, vol. 51, no. 1, pp. 386-398,
//! Jan. 2005.*
//!
//! The algorithm works by adding edge by edge to the Tanner graph. For each
//! symbol node, `wc` check nodes are selected to be joined by edges. Each one
//! is selected in a different step, and the edge is added to the graph, which
//! affects subsequent decissions.
//!
//! To select an edge for the current symbol node, a breadth-first search is
//! done with that node as the root, in order to find the distance from each of
//! check nodes to the root. If there are any check nodes not yet reachable from
//! the root, a node at random is selected among the unreachable nodes that
//! have minimum degree (note that this always happens whenever the first edge
//! is added to a symbol node). If all the check nodes are rechable from the
//! root, the set of nodes of minimum degree among those nodes at maximum
//! distance from the root is selected. A node is picked at random from that
//! set.
//!
//! This procedure tries to maximize local girth greedily and to fill the
//! check nodes uniformly.
use crate::rand::{Rng, *};
use crate::sparse::{Node, SparseMatrix};
use crate::util::{compare_some, *};
use std::cmp::Ordering;
use std::fmt;
use std::fmt::{Display, Formatter};
/// Runtime errors of the PEG construction.
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub enum Error {
/// Not enought rows available.
NoAvailRows,
}
impl Display for Error {
fn fmt(&self, f: &mut Formatter<'_>) -> fmt::Result {
match self {
Error::NoAvailRows => write!(f, "not enough rows available"),
}
}
}
impl std::error::Error for Error {}
/// Result type used to indicate PEG runtime errors.
pub type Result<T> = std::result::Result<T, Error>;
/// Configuration for the Progressive Edge Growth construction
///
/// This configuration is used to set the parameters of the
/// LDPC code to construct.
#[derive(Debug, Clone, PartialEq, Eq)]
pub struct Config {
/// Number of rows of the parity check matrix.
pub nrows: usize,
/// Number of columns of the parity check matrix.
pub ncols: usize,
/// Column weight of the parity check matrix.
pub wc: usize,
}
impl Config {
/// Runs the Progressive Edge Growth algorith using a random seed `seed`.
pub fn run(&self, seed: u64) -> Result<SparseMatrix> {
Peg::new(self, seed).run()
}
}
struct Peg {
wc: usize,
h: SparseMatrix,
rng: Rng,
}
impl Peg {
fn new(conf: &Config, seed: u64) -> Peg {
Peg {
wc: conf.wc,
h: SparseMatrix::new(conf.nrows, conf.ncols),
rng: Rng::seed_from_u64(seed),
}
}
fn insert_edge(&mut self, col: usize) -> Result<()> {
let row_dist = self.h.bfs(Node::Col(col)).row_nodes_distance;
let row_num_dist_and_weight: Vec<_> = row_dist
.into_iter()
.enumerate()
.map(|(j, d)| (j, d, self.h.row_weight(j)))
.collect();
let selected_row = row_num_dist_and_weight
.sort_by_random_min(
|(_, x, w), (_, y, v)| match compare_some(x, y).reverse() {
Ordering::Equal => w.cmp(v),
c => c,
},
&mut self.rng,
)
.ok_or(Error::NoAvailRows)?
.0;
self.h.insert(selected_row, col);
Ok(())
}
fn run(mut self) -> Result<SparseMatrix> {
for col in 0..self.h.num_cols() {
for _ in 0..self.wc {
self.insert_edge(col)?;
}
}
Ok(self.h)
}
}