is presented in this paper based on the Min-Sum Algorithms. Mapping network
Address Generator and pipe line design are adopted to reduce the complexity of hardware implementation
decrease the quantity of memory usage
and enhance the throughout of system. The rate thus becomes reconfigurable through sharing resources without increasing any memory usage. Based on this structure
a LDPC decoder for CMMB standard is implemented on FPGA. The throughout can be up to 70.5Mb/s while the rate is 1/2 and iteration is 10
and 73.2Mb/s while the rate is 3/4 and iteration is 15.