Abstract
Joint channel estimation (CE) and multi-user detection (MUD) have become a crucial part of iterative receivers. In this paper, we propose a quantum-assisted repeated weighted boosting search (QRWBS) algorithm for CE and we employ it in the uplink of multiple-input multiple-output orthogonal frequency division multiplexing systems, in conjunction with the maximum a posteriori probability (MAP) MUD and a near-optimal quantum-assisted MUD (QMUD). The performance of the QRWBS-aided CE is evaluated in rank-deficient systems, where the number of receive antenna elements (AEs) at the base station (BS) is lower than the number of supported users. The effect of the channel impulse response prediction filters, of the power delay profile of the channels, and of the Doppler frequency on the attainable system performance is also quantified. The proposed QRWBS-aided CE is shown to outperform the RWBS-aided CE, despite requiring a lower complexity, in systems where iterations are invoked between the MUD, the CE, and the channel decoders at the receiver. In a system, where $U=7$ users are supported with the aid of $P=4$ receive AEs, the joint QRWBS-aided CE and QMUD achieves a 2-dB gain, when compared with the joint RWBS-aided CE and MAP MUD, despite imposing 43% lower complexity.
Highlights
In the uplink of high-velocity multi-user, multi-carrier systems, the complexity imposed by accurately estimating the channels, as well as detecting the transmitted symbols may become excessive
In our previous work we have proposed quantum-assisted algorithms for providing near-optimal hard-input hard-output (HIHO) Quantumasissted MultiUser Detection (MUD) (QMUD) [7], [29], as well as soft-input soft-output (SISO) quantum-assisted MUD (QMUD) [7], [30]–[32], which may be employed in iterative receivers and were found to be superior both to the conventional Zero-Forcing (ZF) and Minimum Mean Square Error (MMSE) detectors, as well as to the Ant Colony Optimization (ACO) [30]
We show our QRWBSaided Channel Estimation (CE) achieves a better performance than the classic RWBS-aided CE, despite its lower complexity
Summary
In the uplink of high-velocity multi-user, multi-carrier systems, the complexity imposed by accurately estimating the channels, as well as detecting the transmitted symbols may become excessive. Various techniques have been proposed for providing Channel Estimation (CE) with the aid of pilot training symbols [3], [4] as well as low-complexity MultiUser Detection (MUD) [2], [5]–[7]. In [19], Zhang et al employed discrete-space and continuous-space evolutionary algorithms in the MUD and the CE, respectively, for performing joint channel estimation and multi-user detection. Quantum computing [21]–[23] may support the process of joint CE and MUD by exploiting its inherent parallelism for reducing the complexity and for improving the data detection’s and channel estimation’s performance. We compare it to a system, where either the optimal Maximum A posteriori Probability (MAP) MUD or the RWBS-aided CE are employed, demonstrating that the quantum-assisted joint CE and MUD achieve both a better performance and lower complexity.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
