A source-channel combined coding method including: determining whether a channel signal-to-noise ratio (SNR) is varied or not; when it is determined that the channel SNR is varied, selecting a MODCOD suitable for the channel SNR by referring to a first table defining an SNR threshold value at which data transmission is performed without an error, for each MODCOD designating a low density parity check (LDPC) code rate and a modulation scheme; calculating a source coding rate by using an effective information bit rate of the selected MODCOD; extracting network abstraction layer (NAL) units for each layer from an inputted video frame so as to satisfy the calculated source coding rate, and packetizing the extracted NAL units; binding packets to configure a baseband (BB) frame; and LDPC coding and modulating the BB frame through the code rate and the modulation scheme which are designated by the selected MODCOD.