CSI Phase Research Articles

Device-Free Wireless Sensing for Gesture Recognition Based on Complementary CSI Amplitude and Phase.

By integrating sensing capability into wireless communication, wireless sensing technology has become a promising contactless and non-line-of-sight sensing paradigm to explore the dynamic characteristics of channel state information (CSI) for recognizing human behaviors. In this paper, we develop an effective device-free human gesture recognition (HGR) system based on WiFi wireless sensing technology in which the complementary CSI amplitude and phase of communication link are jointly exploited. To improve the quality of collected CSI, a linear transform-based data processing method is first used to eliminate the phase offset and noise and to reduce the impact of multi-path effects. Then, six different time and frequency domain features are chosen for both amplitude and phase, including the mean, variance, root mean square, interquartile range, energy entropy and power spectral entropy, and a feature selection algorithm to remove irrelevant and redundant features is proposed based on filtering and principal component analysis methods, resulting in the construction of a feature subspace to distinguish different gestures. On this basis, a support vector machine-based stacking algorithm is proposed for gesture classification based on the selected and complementary amplitude and phase features. Lastly, we conduct experiments under a practical scenario with one transmitter and receiver. The results demonstrate that the average accuracy of the proposed HGR system is 98.3% and that the F1-score is over 97%.

WiPhase: A Human Activity Recognition Approach by Fusing of Reconstructed WiFi CSI Phase Features

WiPhase: A Human Activity Recognition Approach by Fusing of Reconstructed WiFi CSI Phase Features

Enabling WiFi Sensing on New-generation WiFi Cards

The last few years have witnessed the rapid development of WiFi sensing with a large spectrum of applications enabled. However, existing works mainly leverage the obsolete 802.11n WiFi cards (i.e., Intel 5300 and Atheros AR9k series cards) for sensing. On the other hand, the mainstream WiFi protocols currently in use are 802.11ac/ax and commodity WiFi products on the market are equipped with new-generation WiFi chips such as Broadcom BCM43794 and Qualcomm QCN5054. After conducting some benchmark experiments, we find that WiFi sensing has problems working on these new cards. The new communication features (e.g., MU-MIMO) designed to facilitate data transmissions negatively impact WiFi sensing. Conventional CSI base signals such as CSI amplitude and/or CSI phase difference between antennas which worked well on Intel 5300 802.11n WiFi card may fail on new cards. In this paper, we propose delicate signal processing schemes to make wireless sensing work well on these new WiFi cards. We employ two typical sensing applications, i.e., human respiration monitoring and human trajectory tracking to demonstrate the effectiveness of the proposed schemes. We believe it is critical to ensure WiFi sensing compatible with the latest WiFi protocols and this work moves one important step towards real-life adoption of WiFi sensing.

Simple Solution Preparation of Cs2SnI6 Films and Their Applications in Solid-State DSSCs.

Cs2SnI6 powder is, for the first time, solution-prepared via the formula CsI + SnI2 + I2 → Cs2SnI6. The product is highly pure and air/thermal stable. It is found that N,N-dimethylformamide (DMF) and methanol induce severe Cs2SnI6 deterioration with the appearance of a CsI phase in film preparation from Cs2SnI6 powder, while solvents of γ-butyrolactone (GBL) and ethylene glycol methyl ether (EGME) (Film-EGME) give better results. Then, by introducing EGME solvent, in situ preparation of Cs2SnI6 films (Film-1 to Film-4) is realized under solution reaction, which is found to be dominated by thermal dynamic process, i.e., highly pure/oriented Film-4 is obtained under the maximum reagent-concentration. Besides, for good reaction, the solubility of solvent should be balanced among all the reagents and products. Solid-state dye sensitized solar cells (ss-DSSCs) comprising a Cs2SnI6 electrolyte are investigated. The power conversion efficiencies (PCEs) of the ss-DSSCs based on solution-casted Film-EGME and the in situ-prepared Film-4 are 1.81% and 3.30%, respectively. Particularly, with the in situ prepared Cs2SnI6 films, it is found that the open circuit voltages of the ss-DSSCs are closely related to their gap states. When additive is added in Cs2SnI6 electrolyte, a PCE of 6.14% is obtained in an ss-DSSC. Our work highlights the importance of solvent in film preparation and the role of Cs2SnI6 gap states in device performance.

Synthesis and Characterization of High-Efficiency Halide Perovskite Nanomaterials for Light-Absorbing Applications.

Inorganic perovskite materials are possible candidates for conversion of solar energy to electrical energy due to their high absorption coefficient. Perovskite solar cells (PSCs) introduced a new type of device structure that has attention due to better efficiencies and interest in PSCs that has been increasing in recent years. Halide perovskite materials such as CsPbIBr2 show remarkable optical and structural performance with their better physical properties. Perovskite solar cells are a possible candidate to replace conventional silicon solar panels. In the present study, CsPbIBr2 perovskite materials' thin films were prepared for light-absorbing application. Five thin films were deposited on the glass substrates by subsequent spin-coating of CsI and PbBr2 solutions, subsequently annealed at different temperature values (as-deposited, 100, 150, 200 and 250 °C) to get CsPbIBr2 thin films with a better crystal structure. Structural characterizations were made by using X-ray diffraction. CsPbIBr2 thin films were found to be polycrystalline in nature. With increasing annealing temperature, the crystallinity was improved, and the crystalline size was increased. Optical properties were studied by using transmission data, and by increasing annealing temperature, a small variation in optical band gap energy was observed in the range of 1.70-1.83 eV. The conductivity of CsPbIBr2 thin films was determined by a hot probe technique and was found to have little fluctuating response toward p-type conductivity, which may be due to intrinsic defects or presence of CsI phase, but a stable intrinsic nature was observed. The obtained physical properties of CsPbIBr2 thin films suggest them as a suitable candidate as a light-harvesting layer. These thin films could be an especially good partner with Si or other lower band gap energy materials in tandem solar cells (TSC). CsPbIBr2 material will harvest light having energy of ∼1.7 eV or higher, while a lower energy part of the solar spectrum will be absorbed in the partner part of the TSC.

Deep Learning Phase Compression for MIMO CSI Feedback by Exploiting FDD Channel Reciprocity

Large scale MIMO FDD systems are often hampered by bandwidth required to feedback downlink CSI. Previous works have made notable progresses in efficient CSI encoding and recovery by taking advantage of FDD uplink/downlink reciprocity between their CSI magnitudes. Such framework separately encodes CSI phase and magnitude. To further enhance feedback efficiency, we propose a new deep learning architecture for phase encoding based on limited CSI feedback and magnitude-aided information. Our contribution features a framework with a modified loss function to enable end-to-end joint optimization of CSI magnitude and phase recovery. Our test results show superior performance in indoor/outdoor scenarios.

The scintillation performance of one-inch diameter CsI/CsCl/NaCl eutectics grown by the Czochralski method

In this study, one-inch diameter Tl:CsI/CsCl/NaCl eutectic scintillators were grown by the Czochralski method at various pulling rates. Crystal growth conditions such as rotation, pulling rates, and crystal growth setup were analyzed. By examining the structure of the obtained eutectic, the growth orientation of the CsI phase as well as the optimum pulling speed and rotation speed were found. A transparent eutectic plate of about 3 × 5 mm was prepared from the grown sample, and scintillation properties such as light output and decay time after excitation by 662 keV gamma-rays were evaluated. The performance was compared with the Tl: CsI single crystal standard. Furthermore, a β-ray imaging test was carried out by combining the eutectic plate and an electron-multiplied charge coupling device camera, and the position resolution as an imaging device was evaluated.

Growth and scintillation properties of Tl-doped CsI/CsCl/NaCl ternary eutectic scintillators

Tl:CsI/CsCl/NaCl eutectic scintillators were grown by the Bridgman–Stockbarger method. The eutectic phases were uniformly distributed in the transverse direction and slightly aligned along the growth direction. Then, the CsI phase can be designed as columnar. The expected emission peak observed at 550 nm was observed under X-ray excitation and cathodoluminescence and was ascribed to the self-trapped exciton perturbed by TI+ in Tl:CsI. The mapping of optically waveguided light emission was observed by confocal laser microscopy, consequently optical waveguiding of Tl+ emission in the CsI phase was demonstrated.

Deep Learning Sign Language Recognition System Based on Wi-Fi CSI

Many sensing gesture recognition systems based on Wi-Fi signals are introduced because of the commercial off-the-shelf Wi-Fi devices without any need for additional equipment. In this paper, a deep learning-based sign language recognition system is proposed. Wi-Fi CSI amplitude and phase information is used as input to the proposed model. The proposed model uses three types of deep learning: CNN, LSTM, and ABLSTM with a complete study of the impact of optimizers, the use of amplitude and phase of CSI, and preprocessing phase. Accuracy, F-score, Precision, and recall are used as performance metrics to evaluate the proposed model. The proposed model achieves 99.855%, 99.674%, 99.734%, and 93.84% average recognition accuracy for the lab, home, lab + home, and 5 different users in a lab environment, respectively. Experimental results show that the proposed model can effectively detect sign gestures in complex environments compared with some deep learning recognition models.

DFPhaseFL: a robust device-free passive fingerprinting wireless localization system using CSI phase information

Device-free passive wireless indoor localization is attracting great interest in recent years due to the widespread deployment of Wi-Fi devices and the numerous location-based services requirements. In this paper, we propose DFPhaseFL, the first device-free fingerprinting indoor localization system that purely uses CSI phase information. It utilizes the CSI phase information extracted from simply a single link to estimate the location of the target, neither requiring the target to wear any electronic equipment nor deploying a large number of access points and monitor devices. In DFPhaseFL, the raw CSI phases are extracted from the CSI measurements through the three antennas of the Intel WiFi Link 5300 wireless Network Interface Card (IWL 5300 NIC) firstly. Then, linear transformation and noise filtering are applied to acquire the calibrated CSI phases. Through experimental observations, we find that the calibrated CSI phase owns an unpredictable characteristic over time. Thus, it cannot be directly applied as a fingerprint. To this end, a transfer deep supervised neural network method combining deep neural network and transfer learning is proposed to obtain feature representations with both transferability and discriminability as fingerprints. Then, the DFPhaseFL system uses the SVM algorithm to obtain the estimation of the target location online. Experiment results demonstrate that the DFPhaseFL owns a better estimation precision compared with the other state of art, and maintain a stable localization accuracy for a long time without reacquiring the fingerprint database.

FingerDraw

This paper explores the possibility of tracking finger drawings in the air leveraging WiFi signals from commodity devices. Prior solutions typically require user to hold a wireless transmitter, or need proprietary wireless hardware. They can only recognize a small set of pre-defined hand gestures. This paper introduces FingerDraw, the first sub-wavelength level finger motion tracking system using commodity WiFi devices, without attaching any sensor to finger. FingerDraw can reconstruct finger drawing trajectory such as digits, alphabets, and symbols with the setting of one WiFi transmitter and two WiFi receivers. It uses a two-antenna receiver to sense the sub-wavelength scale displacement of finger motion in each direction. The theoretical underpinning of FingerDraw is our proposed CSI-quotient model, which uses the channel quotient between two antennas of the receiver to cancel out the noise in CSI amplitude and the random offsets in CSI phase, and quantifies the correlation between CSI value dynamics and object displacement. This channel quotient is sensitive to and enables us to detect small changes in In-phase and Quadrature parts of channel state information due to finger movement. Our experimental results show that the overall median tracking accuracy is 1.27 cm, and the recognition of drawing ten digits in the air achieves an average accuracy of over 93.0%.

Vapor assisted deposition of alkaline doped perovskites: Pure phase formation of CsxMA1−xPbI3

Alkaline doping (Cs, Rb) in CH3NH3PbI3 (MAPbI3) is known to enhance the stability of perovskite solar cells. The films are usually deposited using anti-solvent method, which is tricky and not applicable for large scale deposition. Besides, in case of CsxMA1−xPbI3 the amount of Cs must be carefully controlled to prevent CsI phase formation. Herein, we report an atmospheric pressure vapor assisted solution process (AP-VASP) for the growth of Cs doped MAPbI3 perovskite films that features highly uniform morphology, pin-hole free films, large grain size, as well as being scalable. The CsxMA1−xPbI3 films are formed by the reaction of Cs doped PbI2 films with MAI vapor in a simple oven. We demonstrate a simple method for obtaining CsxMA1−xPbI3 films with Cs content as high as 20% without phase segregation of CsPbI3 or CsI. Impedance spectra measurements reveal that Cs incorporation into MAPbI3 increases recombination resistance, up to 83% (28 vs. 15 Ω cm2), resulting in higher open circuit voltage and efficiency. The photovoltaic performance is considerably improved by Cs doping, with the highest efficiency being 14.1% for the Cs0.1MA0.9PbI3 cells compared with 13% for the MAPbI3 ones. As expected, Cs0.1MA0.9PbI3 cells show better stability compared to MAPbI3 cells.

In situ optical absorption spectroscopy of annealing behaviours ofquench-deposited films in the binary system CsI– PbI2

We have investigated annealing behaviours of quench-deposited films of the binary CsI–PbI2system by in situ optical absorption spectroscopy. Variousfilms composed of multiple crystalline phases of CsPbI3 and/or Cs4PbI6 as well as ofCsI and/or PbI2are obtained, depending on the mixing ratio of the CsI and PbI2.It is difficult to prepare films purely composed of a single CsPbI3 or Cs4PbI6phase alone. However, it is possible to obtain films where crystallites of either CsPbI3 or Cs4PbI6coexist with the CsI phase. Using such films, we measurethe fundamental optical absorption spectrum of CsPbI3 and Cs4PbI6 for the firsttime. Cs4PbI6exhibits stronger oscillator-like absorption compared to CsPbI3,due to the localized nature of both the Pb 6s and 6p states.