| Intelligent reflecting surface(IRS)can significantly improve the energy efficiency,spectrum efficiency,and coverage performance of wireless communication networks in a costeffective and energy-saving manner,making it a highly competitive and revolutionary technology for future wireless networks and one of the alternative technologies for 6G communications.Ultra-low latency is one of the key performance indicators of 5G/6G communications,and there are a large number of latency-sensitive applications in the ndustrial Internet of Things and the Vehicle-to-Everything.Most of the existing studies,however,only focus on the improvement of performance indicators such as the power efficiency,spectral efficiency and coverage of wireless networks due to the application of IRSs,ignoring the end-to-end communication delay caused by it.To tackle this issue,this dissertation conducts research on the IRS-assisted wireless transmission technologies including reflection modulation,channel estimation and passive beamforming designs for delay-sensitive applications.First,a reflection pattern modulation(RPM)technique,which encodes the IRS information in the ON/OFF states of the IRS reflecting elements,is proposed by leveraging the concept of index modulation.For the scenario where the access point(AP)and IRS are two independent information sources,this dissertation formulates an optimization problem to minimize the outage probability by jointly optimizing the active beamforming at the AP and passive beamforming at the IRS under the assumption that the IRS state information is statistically known by the AP.Under the SISO system setting and Rician fading channel model,this dissertation derives a closed-form expression for the asymptotic outage probability of the RPM system with respect to the signal-to-noise ratio.By assuming that the AP is sending discrete constellation modulation signals,this dissertation analyzes the achievable rate of the RPM system,and reveals the trade-off between the outage probability and achievable rate via Monte Carlo simulations.Since the RPM scheme incurs a degradation in the IRS reflection power,a quadrature reflection pattern modulation(QRPM)scheme is further proposed to encode the IRS information in the phase shifts of the IRS reflecting elements.Under the SISO system setting and the Rician fading channel model,this dissertation derives a closed-form expression for the pairwise error probability and the upper bound on the average bit error probability of the QRPM system with the maximum likelihood detection.Moreover,this dissertation further proposes a low-complexity demodulation method by first demodulating the constellation symbols transmitted by the AP and then the symbols modulated by the IRS successively.Monte Carlo simulation results show that the QRPM scheme improves the received signal power and thus copes with noise better,as compared to the RPM scheme.Then,this dissertation studies practical design issues in IRS-assisted wireless system including transmission frame structure,channel state information(CSI)acquisition,and passive beamforming optimization,for delay-sensitive applications.Considering that in practice wireless networks usually operate in broadband channels with frequency selectivity and the passive beamforming optimization problem in IRS-assisted multi-carrier systems is very complicated,this dissertation considers the IRS-assisted broadband OFDM system.To tackle the end-to-end communication delay issue caused by acquiring all the cascaded channel coefficients at one time,this dissertation proposes a new frame structure to execute channel estimation and data transmission in a alternating and periodical manner.Furthermore,as the comb-type pilot structure is considered,it is proposed to design the IRS reflection state during channel estimation by superimposing the progressively updated passive beamforming on the columns of the IRS’s reflection pattern,so as to enhance the channel gains on the data tones of training OFDM symbols,thereby increasing the average transmission rate.Based on the estimated CSI of different granularities at the AP,by assuming the continuous phase-shifting model at each IRS element,this dissertation proposes a low-complexity passive beamforming design for maximizing the average achievable rate.Since the formulated problem is non-convex,an alternate optimization(AO)algorithm based on the semi-definite relaxation(SDR)method is proposed to obtain a suboptimal solution to it.However,the computational complexity of the SDR-based AO algorithm is relatively high as it generally requires a large number of iterations.Hence,a low-complexity optimization algorithm is further proposed to solve the formulated problem effectively by exploiting the fact that the channel energy in the time domain is more concentrated than that in the frequency domain.For the scenario with line-of-sight(Lo S)dominant channels,the proposed low-complexity optimization algorithm can achieve near-optimal performance as well. |
