This paper considers signal-to-noise ratio (SNR) estimation in transmission systems with an iteratively decodable low-density parity-check (LDPC) code by means of the expectation-maximization (EM) algorithm. We present a simplified version of the code-aided (CA) estimation algorithm by introducing the embedded estimation technique to reduce the computational complexity related to the iterative process of the estimator. And we also show that the conventional non-data-aided (NDA) estimators can be interpreted as a special case of the CA estimator. Computer simulations show that the improve CA approach can achieve good estimates at low SNR with significantly less computation.