LDPC Validation Service

One of the most effective forms of forward error correction (FEC) codes used in today’s commercial and military communication systems is Low-Density Parity-Check (LDPC) codes.  These codes offer excellent Bit Error Rate (BER) and Block Error Rate (BLER) performance, with capacity approaching the Shannon limit.  LDPC codes are used in many standards including DBV-S2, 802.16, IEEE 802.11n, IEEE 802.11ad, and DOCSIS 3.1.

Avaliant provides a high-throughput, flexible, and cost-effective LDPC Validation service.  The service allows the user to measure and thus determine the performance of LDPC codes with astonishing precision. This enables clients to calculate the benefit of implementing specific LDPC codes in their system. LDPC Validator has a rapid turnaround time because it was designed to seamlessly take advantage of a massively parallel General Purpose Graphics Processor (GPU) architecture. 

The customer can leverage both the speed and flexibility of the Avaliant LDPC Validation service while designing new code, or optimizing existing communication systems before committing to a final design.  As an example, the Figure below shows the rate ½ parity check matrix (H matrix) for IEEE 802.11ad standard.  The H matrix is very sparse with the 1’s represented in red.

 LDPC Validation Service

The LDPC Validation service currently supports the following set of LDPC code parameters:

LDPC Parity Matrix (H):

  • Regular: An LDPC code is called regular if the number of 1’s for every column of H matrix is constant, and the number of 1’s for every row of H matrix is also a constant.
  • Irregular: An LDPC code is called irregular if the number of 1’s in each row or column of H matrix aren’t constant.
  • Punctured: Puncturing is the process of removing some of the parity bits after LDPC encoding.  The removed parity bits are not transmitted over the channel.  Puncturing has the effect of degrading the bit error rate performance of the communications system but also results in increasing the code rate.
  • Shortened:  Shortening is the process of using less data bits than possible as compared to the native LDPC code.  Shortening has the effect of improving the bit error rate performance of the communication system but also results in decreasing the code rate.
  • Varying block size: The size of the H matrix can be as large as 64,000 bits.

LDPC Generator Matrix (G): The G matrix is used in the model to create all possible codewords

LDPC Decoder Parameters:

  • Number of iterations:  Maximum number of times the same codeword is processed before deciding on what data was transmitted
  •  Log-Likelihood Ratio (LLR) definition:  Exact or approximate.  High-order modulations above QPSK can have complex LLRs which can be approximated for computational efficiency

Modulation (DOCSIS 3.1):  BPSK, QPSK, 8QAM, 16QAM, 32QAM, and 64QAM. 

The Avaliant LDPC validation service outputs bit error rate, block error rate, and average iterations per signal-to-noise ratio (SNR).  The service has been validated using known LDPC standards including 802.11ad, 802.16, and DVB-S2:  see ‘Case Study Performance Results Validation’ document for details.  The Avaliant LDPC Validation tool is currently being used by customers such as a large aerospace company.

The interface process between Avaliant and the customer is illustrated in the Table below.

 LDPC Validation Service

Contact:   Bob Bagshaw

Email:   ldpc@avaliant.com

Phone:  425-644-7210 extension 21