Abstract
In this study, a class of Generalized Low-Density Parity-Check (GLDPC) codes is designed for data transmission over a Partial-Band Jamming (PBJ) environment. The GLDPC codes are constructed by replacing parity-check code constraints with those of nonsystematic Bose-Chaudhuri-Hocquenghem (BCH), referred to as Low-Density Parity-Check (LDPC)-BCH codes. The rate of an LDPC-BCH code is adjusted by selecting the transmission length of the nonsystematic BCH code, and a low-complexity decoding algorithm based on message-passing is presented that employs A Posteriori Probability (APP) fast BCH transform for decoding the BCH check nodes at each decoding iteration. Simulation results show that the LDPC-BCH codes with a code rate of 1/8.5 have a bit error rate performance of $1 \times 10^{-6}$ at signal-noise-ratios of -6.97 dB, -4.63 dB, and 2.48 dB when the fractions of the band jammed are 30%, 50%, and 70%, respectively.
Introduction
Frequency-Hopping Spread-Spectrum (FH-SS) is a standard defense against interference in wireless communications. However, it remains highly vulnerable to Partial-Band Jamming (PBJ), where an adversary concentrates power on specific frequency slots. While standard Low-Density Parity-Check (LDPC) codes are effective in Gaussian noise, they struggle in PBJ scenarios where bursty interference can overwhelm simple parity checks. This paper introduces LDPC-BCH codes, a novel class of Generalized LDPC (GLDPC) codes designed to survive extreme interference by replacing standard check nodes with more powerful error-correcting components.
System Architecture and Code Construction
The proposed communication system integrates the LDPC-BCH encoder and decoder into a frequency-hopping framework designed to mitigate PBJ effects. The overall system model, including the frequency hopper and the PBJ channel, is shown in Figure 1.
The core innovation is the structural generalization of the LDPC graph. In a typical LDPC code, "check nodes" perform simple parity checks. The authors replaced these with nonsystematic Bose-Chaudhuri-Hocquenghem (BCH) codes. This embeds smaller, powerful error-correcting codes within the larger LDPC graph, significantly increasing resilience to the burst errors caused by jammed frequency hops. Figure 3 illustrates the structural difference between a standard LDP code and the proposed LDPC-BCH code.
Decoding and Performance
To manage the increased computational load of GLDPC decoding, the researchers implemented a message-passing algorithm using a Fast BCH Transform (FBT). This transform uses a "butterfly" flow graph to efficiently calculate A Posteriori Probabilities (APP) at each check node, keeping decoding complexity manageable.
Simulation results demonstrate the robustness of the LDPC-BCH scheme. The system achieved a Bit Error Rate (BER) of $1 \times 10^{-6}$ even when up to 70% of the communication band was compromised. The required Signal-to-Noise Ratios (SNR) to reach this performance were -6.97 dB (30% jamming), -4.63 dB (50% jamming), and 2.48 dB (70% jamming). This performance significantly outperforms traditional Reed-Solomon-BCH concatenated codes, which often hit an error floor in high-jamming environments.
Conclusion
By generalizing the LDPC structure with nonsystematic BCH check nodes, the proposed LDPC-BCH codes provide a powerful defense against partial-band jamming. The use of a Fast BCH Transform ensures that this increased robustness does not come at the cost of prohibitive computational complexity, making it a viable candidate for secure, high-reliability wireless links.
Summary updated