Cooperative multi‐cellular large MIMO over asynchronous channel training

Abstract

SUMMARYThis work analyzes multi-cell MIMO systems equipped with a very large number of antennas in the basestation (BS) side and with low-complexity zero-forcing and matched-filter receivers. Cooperation between BSs in a multi-cell scenarios is analyzed in order to avoid the well-known pilot contamination effect on massive MIMO systems, while the effects of asynchronism between users and BSs in training phase using different pilot sequences, for example, Walsh-Hadamard, Gold, and Kasami sequence sets on the system performance are examined. As low-complexity processing is deployed, the synchronism offset in training phase may introduce errors on the channel estimation process, adding its effect to the pilot contamination, resulting in an aggressive detection scenario. Numerical results using the BER as a performance figure of merit measure in downlink have corroborated our finding. It was also developed an analytical error formula for channel estimation subject to synchronization offset levels in terms of fractional unit of symbol time duration, which was validated through channel mean squared error and BER performance simulation results. Through these both performance figures and under practical scenarios, the use of pseudo-random sequence sets may be preferable instead of orthogonal ones, mainly because of their out-of-phase cross-correlation levels.

Fairness‐aware nonlinear joint transceiver design for energy‐harvesting‐powered CoMP systems

“…𝜅 = 0, these two algorithms have almost the same performance. From Equation (21), it can be seen that p min effects the transmit power p k l (∀k ∈ , l ∈  k ). With the increasing of 𝜅, p min becomes the dominate element to the transmit power, and decreases the system performance.…”

“…However, it has not been considered in existing schemes. Besides, CSI acquired at nodes were also supposed to be perfect, which is rather unrealistic, as errors would be inevitably introduced by channel estimation, channel quantization or feedback delay in practice [21][22]. The above facts motivate this paper to study on the fairnessaware nonlinear transceiver design under the per-node power constraints, and further on the robust scheme with imperfect CSI for EH-powered CoMP systems.…”

Asynchronous Channel Training in Massive MIMO Systems

“…Without loss of generality, it is assumed that 0 = τ 11 11 < τ 12 11 < τ 21 11 < τ 22 11 < T where T is the symbol duration. For simplicity of illustration, the transmitted signals are depicted with a rectangular pulse shape in Fig.…”

“…We intentionally create timing mismatch between different received signals. While asynchronous channel training method in [11] has resulted in erroneous estimated channel because of noise and channel correlation, we show how to avoid the drawbacks by designing an appropriate sampling method. It has been shown that the performance of MIMO systems with intentional timing offset between transmitters can be improved by using sampling diversity [12][13][14][15].…”

A lower bound on achievable rate of MRT precoding in multicell multiuser massive MIMO networks with Rician flat fading

“…By their network model, these works can be divided into two main groups: cooperative and noncooperative. The former assumes that base stations of different cells, in order to increase the total capacity, share their information like channel state information or users data [1][2][3][4][5][6][7][8][9][10]. For example, in [3], sum-rate for the multicell vector perturbation precoding in the case of uniformly distributed input and an asymptotic upper bound on the sum-rate at high signal-to-noise ratios are derived.…”