Enhancing the Error-Correcting Performance of LDPC Codes through an Efficient Use of Decoding Iterations

Authors: Insah Bhurtah, P. Clarel Catherine, K. M. Sunjiv Soyjaudah

Abstract:

The decoding of Low-Density Parity-Check (LDPC) codes is operated over a redundant structure known as the bipartite graph, meaning that the full set of bit nodes is not absolutely necessary for decoder convergence. In 2008, Soyjaudah and Catherine designed a recovery algorithm for LDPC codes based on this assumption and showed that the error-correcting performance of their codes outperformed conventional LDPC Codes. In this work, the use of the recovery algorithm is further explored to test the performance of LDPC codes while the number of iterations is progressively increased. For experiments conducted with small blocklengths of up to 800 bits and number of iterations of up to 2000, the results interestingly demonstrate that contrary to conventional wisdom, the error-correcting performance keeps increasing with increasing number of iterations.

Keywords: Error-correcting codes, information theory, low-density parity-check codes, sum-product algorithm.

Digital Object Identifier (DOI): doi.org/10.5281/zenodo.1328890

References:

[1]R. Gallager, ""Low-density paarity-check codess,” IRE Transac tions on Information ThTheory, vol. 8, no. 1, pp. 21–28, Jann. 1962.
[2]D. MacKay, "Good error-coorrecting codes based on veryy sparse matrices,” IEEEE Transactions oon Information TTheory, vol. 45, nno. 2, pp. 399–431,Marcch 1999.
[3]T. J. Richards on and R.L. Urb anke, "The capaccity of low-densiity parity check codes uunder message-passing decoding,,” IEEE Transacctions on Information ThTheory, vol. 47, noo. 2, pp. 599–618 , Feb. 2001.
[4]O.H. Kazancii, "Performance of pseudo-randoom and quasi-cyyclic low density parity -check codes,” MMaster’s thesis, The Graduate Scchool of Natural and AApplied Science s of Middle Eaast Technical Unniversity,Turkey, 2007.
[5]K.M.S. Soyjauudah and P.C. Caatherine, "Efficiennt recovery technnique for low-density paarity-check codess using reduced-sset decoding,” Joournal of Circuits, Systeems, and Computters (JCSC), vol.. 17, no. 2, pp. 3333–351, Apr. 2008.
[6]M.C. Davey, "Error-correctionn using low-denssity parity-checkk codes,” Ph.D. dissertattion, University oof Cambridge, Unnited Kingdom, 1 999.
[7]P.C. Catherinne and K. M. S . Soyjaudah, "EErasing bit nodess on the bipartite grapph for enhancedd performance of LDPC codees,” 4th International Conference on Emerging Trennds in Engineerring and Technology (ICICETET), pp. 107 –111, Nov. 2011 .
[8]C.E. Shannon,, "A mathematicaal theory of commmunication,” Belll System Technical Jou rnal, vol. 27, pp. 379–423, 1948.
[9]C. Caini, H. Cruickshank, S. Farrell, and M. Marchese, "De lay- and disruption-toleerant networking (DTN): an alternnative solution fofor future satellite netwoorking applicationns,” Proceedings of the IEEE, voll. 99, no.11, pp. 1980–11997, Nov. 2011..