Pipelined architecture for low density parity check encoder

This paper proposes a pipelined architecture for low density parity check encoder by pipelining information bits and sub-matrices of parity check matrix (H) using two bit-wise operations. The two bit-wise operations are multiplication and exclusive-OR. The investigation is done by exploring two methods of pipelining using the two bit-wise operations in the proposed architecture. The first method of pipelining uses combination of shift register and memory for the sub-matrices of H while the second method of pipelining uses shift register for the information bits and the sub-matrices of H. It is shown that the second method of pipelining increases the throughput but has the largest units of shift registers and flip-flops in the design.