Journal of Information and Intelligence最新文献

Reconfigurable intelligent surfaces (RISs) have aroused extensive attentions from academic and wireless communication communities due to their abilities to customize the electromagnetic (EM) characteristics of the propagation channels flexibly and rapidly. Recent advances in theoretical innovations and prototype systems have demonstrated the advantages of RISs in terms of low cost, low power consumption, and easy deployment. Meanwhile, the optically transparent RISs are demanded in some application scenarios. In this paper, we propose a 2-bit metalmesh-based RIS with high optical-transparency. By analyzing the surface current distributions on the element, we employ the metalmesh-grid patterns and metalmesh-stripe patterns on the top and ground layers respectively. The metalmesh patterns can help improve the optical transparency of RISs, while maintaining similar microwave characteristics. The RIS can reach the optical transparency of 79%, and the reflection amplitude is greater than −3.2 ​dB within the operating band. Finally, to verify the capability of the proposed RIS in wavefront controls, the far-field scattering patterns of the RIS with different coding sequences are investigated and the simulation results are in good agreement with the theoretical results.

In this paper, we proposed an active metasurface in reflection manner that can generate reconfigurable OAM vortex beams with high purity in the X-band. The metasurface has a high reflectance of 0.94 and achieves a phase coverage of 320° between 9.8 ​GHz and 11 ​GHz. Then, by encoding the phase distribution of the meta-atoms, various OAM vortex beams including ±1, ±2, ±3, and ±4 orders are generated, where the purity of all modes can be above 70%. Moreover, the metasurface can also deflect the OAM beam with a certain angle while maintaining high purity, which can be applied to reduce the influence of the alignment deviation between transmitting and receiving antennas during the communication processes. As a validation, the metasurface composed of 30 ​× ​30 meta-atoms is fabricated and measured. Both simulation and measurement results demonstrate the capability of the proposed metasurface to generate reconfigurable OAM beams with high purity, indicating the application potentials of proposed meta-devices in future OAM communications.

This paper proposes a new method to generate a two-dimensional (2D) Airy beam and Airy autofocusing beam by using the scalar holographic metasurface with amplitude-phase modulation in the microwave band. The proposed holographic metasurface comprises subwavelength patch unit cells with a period of fewer than 1/8 wavelengths, which means that it has the finer sampling for electromagnetic waves and can simultaneously achieve precise modulations for the amplitude and phase of electromagnetic waves. Firstly, the 2D-Airy beam with quasi-non-diffraction and self-bending characteristics is generated, from which the holographic metasurface is designed to realize four different 2D-Airy beams with the same focus, achieving the 2D-Airy autofocusing beam in the microwave frequency. The holographic metasurface for Airy beam generation has high efficiency and an ultra-lower profile. Meanwhile, for applying the Airy beam in wireless power transfer (WPT), the efficiency of the generated Airy beam and Airy autofocusing beam is calculated for the first time in the microwave field. The simulation results show that the efficiency of the 2D-Airy beam can reach 66% at 150 ​mm away from the metasurface, while the efficiency of the 2D-Airy autofocusing beam at the focus, which is 280 ​mm from the metasurface, can reach 35%. The theoretical, simulated, and measured results show that the proposed method and holographic metasurfaces can flexibly achieve the special characteristics of self-autofocusing and self-bending Airy beams in the microwave domain, providing an effective path for wireless power transfer (WPT) scenario with radial obstructions.

Reconfigurable intelligent surface (RIS) is emerged as a promising technique to solve the challenges faced by future wireless communication networks. Although the most commonly used electrically-controlled RISs can achieve millisecond-scale speed of dynamic switch, they have a large number of microwave circuit elements (such as PIN diodes or varactors) which will bring non-negligible insertion loss, and the complicity of the bias network to electrically addressing each element will increase with the expansion of the RIS aperture. Aiming at further reducing the fabrication cost and power consumption, herein an electromechanical RIS used for sub-6G wireless communication is proposed. The electromechanical RIS is designed with a passive metasurface and step-motor driver modules, providing simultaneous high-efficiency reflection (over 80%) and continuous reflection phase coverage of 360°. Through electromechanical control, the RIS system can realize different reflective wavefront shaping, and has been employed in the indoor sub-6G wireless environment demonstrating a maximum signal improvement of 8.3 ​dB. The proposed electromechanical RIS is particularly useful for wireless signal enhancement in static blind area, and has the obvious advantage of not requiring continuous power supply after the RIS being regulated. Therefore, it greatly reduces the overall cost and power consumption which may have potentials in indoor application scenarios for improving wireless communication performance.

With the help of a developing technology called reconfigurable intelligent surfaces (RISs), it is possible to modify the propagation environment and boost the data rates of wireless communication networks. In this article, we optimized the phases of the RIS elements and performed a fair power allocation for each subcarrier over the full bandwidth in a single-input-single-output (SISO) wideband system where the user and the access point (AP) are provided with a single antenna. The data rate or its equivalent channel power is maximized by proposing different low-complex algorithms. The strongest tap maximization (STM) and power methods are compared with the semidefinite relaxation (SDR) method in terms of computational complexity and data rate performance. Runtime and complexity analysis of the suggested methods are computed and compared to reveal the actual time consumption and the required number of operations for each method. Simulation results show that with an optimized RIS, the sum rate is 2.5 times higher than with an unconfigured surface, demonstrating the RIS's tremendous advantages even in complex configurations. The data rate performance of the SDR method is higher than the power method and less than the STM method but with higher computational complexity, more than 6 million complex operations, and 50 ​min of runtime calculations compared with the other STM and power optimization methods.

It is of great importance to control flexibly wireless links in the modern society, especially with the advent of the Internet of Things (IoT), fifth-generation communication (5G), and beyond. Recently, we have witnessed that programmable metasurface (PM) or reconfigurable intelligent surface (RIS) has become a key enabling technology for manipulating flexibly the wireless link; however, one fundamental but challenging issue is to online design the PM's control sequence in a complicated wireless environment, such as the real-world indoor environment. Here, we propose a reinforcement learning (RL) approach to online control of the PM and thus in-situ improve the quality of the underline wireless link. We designed an inexpensive one-bit PM working at around 2.442 ​GHz and developed associated RL algorithms, and demonstrated experimentally that it is capable of enhancing the quality of commodity wireless link by a factor of about 10 ​dB and beyond in multiple scenarios, even if the wireless transmitter is in the glancing angle of the PM in the real-world indoor environment. Moreover, we also prove that our RL algorithm can be extended to improve the wireless signals of receivers in dual-receiver scenario. We faithfully expect that the presented technique could hold important potentials in future wireless communication, smart homes, and many other fields.

Reconfigurable intelligent surfaces (RISs) are a promising technology for wireless communication applications, but their performance is often optimized using simplified electromagnetic reradiation models. In this study, we explore the impact on the RIS performance of more realistic assumptions, including the (possibly imperfect) quantization of the reflection coefficients, sub-wavelength inter-element spacing, near-field location, and presence of electromagnetic interference. We find that design constraints can cause an RIS to reradiate power in unwanted directions. Therefore, it is important to optimize an RIS by considering the entire reradiation pattern. Overall, our study indicates that a 2-bit digitally controllable RIS with a nearly constant reflection amplitude and RIS elements with a size and inter-element spacing between (1/8)th and (1/4)th of the signal wavelength may offer a reasonable tradeoff between performance, complexity, and cost.

In view of the difficulty of obtaining downlink channel state information, partial reciprocity based channel covariance matrix (CCM) reconstruction has attracted a lot of attention in frequency division duplex (FDD) multi-antenna systems. Taking both the impact of CCM reconstruction on system performance and design complexity, we investigate an adaptive CCM reconstruction in this paper. Specifically, to effectively evaluate the validity of the reciprocity, we firstly analyze the characteristics of the partial reciprocity and define a reciprocity evaluation criterion. Then, we propose a partial antenna based angular power spectrum (APS) estimating algorithm to further reduce the complexity of the CCM reconstruction. Finally, simulation results demonstrate the superiority of our proposed schemes.

Deep learning based channel state information (CSI) fingerprint indoor localization schemes need to collect massive labeled data samples for training, and the parameters of the deep neural network are used as the fingerprints. However, the indoor environment may change, and the previously constructed fingerprint may not be valid for the changed environment. In order to adapt to the changed environment, it requires to recollect massive amount of labeled data samples and perform the training again, which is labor-intensive and time-consuming. In order to overcome this drawback, in this paper, we propose one novel domain adversarial neural network (DANN) based CSI Fingerprint Indoor Localization (D-Fi) scheme, which only needs the unlabeled data samples from the changed environment to update the fingerprint to adapt to the changed environment. Specifically, the previous environment and changed environment are treated as the source domain and the target domain, respectively. The DANN consists of the classification path and the domain-adversarial path, which share the same feature extractor. In the offline phase, the labeled CSI samples are collected as source domain samples to train the neural network of the classification path, while in the online phase, for the changed environment, only the unlabeled CSI samples are collected as target domain samples to train the neural network of the domain-adversarial path to update parameters of the feature extractor. In this case, the feature extractor extracts the common features from both the source domain samples corresponding to the previous environment and the target domain samples corresponding to the changed environment. Experiment results show that for the changed localization environment, the proposed D-Fi scheme significantly outperforms the existing convolutional neural network (CNN) based scheme.

In this paper, based on the block Markov superposition transmission (BMST) technique, we present a new class of coupled low-density parity-check (LDPC) codes for the transport block (TB)-based transmission to improve the error-correcting performance. For encoding, the previous LDPC codewords corresponding to a TB (at prior time slot) are interleaved and superimposed onto the current LDPC codewords, resulting in the transmitted codewords. For decoding, the sliding window decoding algorithm with sum-product or min-sum implementations can be employed, inheriting a relatively low-latency decoding. A distinguished advantage of the proposed coded transmission over spatially coupled LDPC (SC-LDPC) codes is that the encoder/decoder of the proposed codes can be designed by reusing the encoder/decoder architecture of component block LDPC codes. To analyze the waterfall performance of BMST-LDPC code ensembles, we present the protograph-based EXIT chart analysis, which can efficiently predict the error-correcting performance in waterfall region. To analyze the error-floor performance of BMST-LDPC codes, we employ the genie-aided (GA) lower bound, which can efficiently predict the error-correcting performance in error-floor region. For ease of implementation, the BMST-LDPC codes are constructed by taking the (2, 4)-raptor-like LDPC codes or the 5G LDPC codes as the basic components. The numerical results reveal that the proposed codes can have capacity-approaching performance, exhibiting a gap of 0.007 ​dB away from the corresponding Shannon limit. They also reveal that, by using the proposed BMST construction, the error-correcting performance of the original 5G block LDPC codes can be significantly improved, achieving coding gains up to one dB over the AWGN channels and two dB over the fast fading channels.