This thesis presents a geometric approach to the constructio of low-density parity-check (LDPC) codes. Four Codes of these four A stopping set S in a parity-check matrix H is a subset of the variable nodes in the Tanner graph for H such that all the neighbors of S are connected to S at least twice. Give a parity-check matrix H, the size of the smallest nonempty stopping set is called the stopping distance of H. The stopping redundancy ?(C) of a code C is the minimum number of rows in any arity-check matrix H for C such that s(H) = d(C). It is always possible to find a parity-check matrix H for C such that s(H) = d(C). In fact, the arity-check matrix consisting of all the nonzero codewords of the dual code C? has this property. Unlike the stopping distance, which depends on the ecific choice of H, the stopping redundancy i a property of the code itself. The topping distance and topping redundancy of linear code are important Concept in the analysis of the performance and complexity of the code under iterative decoding o a binary erasure channel. In this thesis, the stopping sets and topping distance of finite geometry LDPC (FG-LDPC) codes are studied. It is hown that the lower bound on the minimum distance of FG-LDPC codes is also a lower bound on the stopping distance of FG-LDPC codes, which implies that this codes have considerably large stopping distance. Thus FG-LDPC codes have a good performance under iterative decoding. Finally topping distance of FG-LDPC codes, constructed based on the points and line of Euclidean or projective geometry, are obtained and a upper bound on the Stopping redundancy of these codes is provided.