Intelligent reflecting surface assisted sparse vector coding based short packet communication for 6G wireless networks

Abstract

Sparse vector coding (SVC) is a viable solution to tackle simultaneously the stringent requirements of achieving high reliability and low latency in short packet communications (SPC). On the other hand, the intelligent reflecting surface (IRS), a cutting-edge technology, shows more significant potential to circumvent obstructions and establish consistent communication using low-cost, energy-efficient reflecting elements. Owing to the advantages of both SVC and IRS, in this work, we integrate IRS and SVC for SPC to connect the unconnected users. We derived the probability density function of the overall cascaded wireless channel, which helped deduce the upper-bound block error rate (BLER) theoretical expressions, and the results suggest improved reliability performance compared to conventional SVC. In addition, we also observed the impact of parameters such as the number of reflecting elements, multiple antennas at the base station, position of the IRS, number of resources, amplitude reflection coefficient, and phase shift coefficient on the IRS-SVC. Further, the proposed model is tested under international telecommunication union (ITU) realistic channels, namely, extended pedestrian-A (EPA), and extended vehicular-A (EVA), and shown to be superior performance over the conventional SVC, with 20.5 dB and 21 dB respectively at BLER point of $10^{-4}$. Lastly, the complexity analysis has been carried out highlighting the trade-off between the reliability and the ease of implementation for the IRS-SVC compared to the conventional SVC.