Simultaneous Channel Estimation and Data Detection Based on Superimposed Training for Many Access MIMO System in Uplink

Abstract

Massive Multiple Input Multiple Output (MIMO) has great potential to improve spectrum efficiency in the fifth generation (5G) wireless communication systems. However, the efficiency was disastrously reduced by the heavy burden of overhead for device detection and channel estimation of the large amount of small data packets in the uplink channel. In the paper, we proposed a novel transmission scheme by superposing the training symbols for active device detection and channel estimation on the data symbols in the uplink transmission to improve the efficiency. More specifically, in order to mitigate the cross interference among the superposed signals, we proposed to split the transmission into the training phase and the traffic phase. Then we superposed the training phase in the next transmission with the traffic phase in the current transmission. Furthermore, we give the optimization of power allocation ratio between training phase and traffic phase to obtain the optimal overall performance. The analytic and simulation results show that, with the help of spatial isolation among devices, our proposed transmission scheme can significantly improve the transmission efficiency compared with the existing schemes.

Optimum Averaging of Superimposed Training Schemes in OFDM Under Realistic Time-Variant Channels

“…In this case, in order to allow a worsening of less than 50% of the minimum MSE, the range of averaging values must befall within [4,16]. Then, in order to compute this optimum value in its allowed range, considering a perfect knowledge of the SNR, the relative speed between the UE and the BS must be v robust 2 [35, 350] km/h, which in terms of Doppler frequency is a range of f robust d 2 [0.9, 9.5] kHz; on the other hand, if there is a perfect knowledge of the Doppler frequency, the SNR range of possible mismatching extends to ⇢ robust SNR 2 [3,22] dB. Obviously, the presented model exhibits a high robustness since the ranges of tolerance to input imprecisions are very broad relative to the resultant MSE computation worsening.…”

“…• opt will not be discussed and the common defined values by academia will be employed ( = 0.2) [6], [11]- [13], [15], [22]. Usually, the optimization of requires further analysis related to the throughput which is not covered in this paper.…”

“…For example, it may be employed when refining the frame synchronization [17], in cooperative relay scenarios [14], as well as in unmanned aerial vehicles (UAV) assisted systems [18]. Adaptation to other waveforms like filter bank multicarrier (FBMC) [19] or to other technologies such as non-orthogonal multipleaccess (NOMA) [20], visible light communications (VLC) [21], MIMO systems with sparse channels [15] or with many access devices [22] is feasible, too.…”

Spectral Efficiency Improvement and Power Control Optimization of Massive MIMO Networks

“…The proposed method considers a decoupled optimization scheme between the RF and baseband domains for the SE maximization problem. In [8], the authors developed a novel transmission scheme by superposing the training symbols for active device detection and channel estimation on the data symbols in the uplink transmission to mitigate the cross interference among the superposed signals and improve the efficiency. In [9], the authors showed that cognitive spectrum sensing and power harvesting can be accomplished simultaneously.…”