Skip to Main Content
This paper presents a data-loading technique that jointly considers the effect of channel estimation and the property of encoded multimedia data in Orthogonal Frequency Division Multiplexing (OFDM) systems. We observe that OFDM subchannels experience different average bit error rate (BER) due to channel estimation inaccuracy. The leakage effect in the fast Fourier transform (FFT)-based channel estimation method or the model mismatch in the polynomial-based channel estimation method results in a variation on the decoded BER across different OFDM subchannels. Thus, we are motivated to design the Priority Transmission (PT) scheme, which utilizes this BER variation across different OFDM subchannels and provides unequal error protection (UEP) for multimedia transmission. In addition, since OFDM has been adopted in many multimedia transmission standards, we compare the different channel estimation techniques, which were compared only for generic data transmission before, in the context of multimedia transmission with the PT scheme. In particular, we extend the polynomial-based channel estimation that was previously designed for a decision-directed scenario to a pilot-symbol-assisted (PSA) channel estimation scenario. Then, we investigate the channel estimation mean square error (MSE) and BER performance of individual OFDM subchannels for both the FFT-based and the polynomial-based channel estimation. Furthermore, we design the PT scheme that achieves significant gain in peak-signal-to-noise ratio (PSNR) of the reconstructed images for both channel estimation methods. Finally, we compare different OFDM channel estimation techniques for multimedia transmission. It is shown that for generic data transmission, the polynomial-based PSA channel estimation outperforms the FFT-based method in realistic channel conditions, and both types of channel estimation have similar performance when using the proposed PT scheme for multimedia transmission.