Performance of decision-feedback channel estimation for TDM based control signals in OFDMA downlink

In orthogonal frequency division multiple access (OFDMA), time division multiplexing (TDM) based layer 1 (L1)/layer 2 (L2) control signals provide advantageous features such as a power saving effect using micro-sleep operation and a short round trip delay (RTD) for channel-dependent scheduling and adaptive modulation and coding (AMC) compared to a frequency division multiplexing (FDM) based structure. However, only a portion of the reference signal (RS) symbols is used for channel estimation assuming staggered RS multiplexing in order to obtain the merits of the TDM based control signals. Hence, this paper presents decision-feedback channel estimation (DFCE) methods using soft-symbol estimation for control signals for the TDM based control signals in the OFDMA downlink. In the proposed DFCE methods, channel estimation accuracy is improved by using soft-symbol estimation of the control signals based on the extrinsic log-likelihood ratio (LLR) at the soft output Viterbi algorithm (SOVA) for convolutional coding in addition to the RS symbols. Computer simulation results show that the required average received signal-to-noise power ratio (SNR) satisfying the average block error rate (BLER) of 10^-2 using DFCE in the subframe of interest is decreased by approximately 0.6 dB compared to that for RS based channel estimation. Hence, we conclude that DFCE with soft-symbol estimation is very effective in improving the channel estimation accuracy using a limited number of RS symbols over the duration in which TDM based control signals are multiplexed, while maintaining the merits such as an efficient power saving effect using micro-sleep operation and a short processing delay.