Multi-access interference(MAI)is the main source limiting the capacity and quality of the multiple-input multipleoutput orthogonal frequency division multiplexing(MIMO-OFDM)system which fulfills the demand of high-spe...Multi-access interference(MAI)is the main source limiting the capacity and quality of the multiple-input multipleoutput orthogonal frequency division multiplexing(MIMO-OFDM)system which fulfills the demand of high-speed transmission rate and high quality of service for future underwater acoustic(UWA)communication.Therefore,multi-user detection(MUD)is needed at the receiver of the MIMO-OFDM system to suppress the effect of MAI.In this research,MUD is achieved by using a criterion based adaptive recursive successive interference cancellation(RSIC)scheme at the receiver of a MIMO-OFDM system whose transceiver model in underwater communication is implemented by using the Bellhop simulation system.The proposed scheme estimates and eliminates the MAI through user signal detection and subtraction from received signals at the receiver of the MIMO-OFDM system in underwater environment.The bit error rate(BER)performance of the proposed scheme is analyzed by using weight filtering and weight selection criteria.By Matlab simulation,it is shown that the BER performance of the proposed scheme outperforms the conventional matched filter(MF)detector,the adaptive successive interference cancellation(SIC)scheme,and the adaptive RSIC scheme in the UWA network.展开更多
In the present paper,MoSi2(Cr5Si3)–RSiC composites were prepared via a combination of precursor impregnation pyrolysis(PIP) and MoSi2-Si-Cr alloy active melt infiltration(AAMI) process. Composition, microstructure, m...In the present paper,MoSi2(Cr5Si3)–RSiC composites were prepared via a combination of precursor impregnation pyrolysis(PIP) and MoSi2-Si-Cr alloy active melt infiltration(AAMI) process. Composition, microstructure, mechanical retention characteristics, and oxidation behaviors of the composites at elevated temperature were studied. X-ray diffraction(XRD) pattern confirms that the composites mainly compose of 6 H–SiC, hexagonal MoSi2, and tetragonal Cr5Si3. Scanning electron microscopy(SEM) image reveals that nearly denseMoSi2(Cr5Si3)–RSiC composites exhibiting three-dimensionally(3D) interpenetrated network structure are obtained when infiltrated at 2173 K, and the interface combination of the composites mainly depends on the composition ratio of infiltrated phases. Oxidation weight gain rate of the composites is much lower than that of RSiC matrix, where MoSiCr2 possesses the lowest value of 0.1630 mg×cm-2, about 78% lower than that of RSiC after oxidation at 1773 K for 100 h. Also, it possesses the highest mechanical values of 139.54 MPa(flexural strength σf and RT) and 276.77 GPa(elastic modulus Ef and RT), improvement of 73.73% and 29.77% as compared with that of RSiC, respectively. Mechanical properties of the composites increase first and then decrease with the extension of oxidation time at 1773 K, due to the cooperation effect of surface defect reduction via oxidation reaction and thermal stress relaxation in the composites, crystal growth, and thickness increase of the oxide film. Fracture toughness of MoSiCr2 reaches 2.24 MPa·m1/2(1673 K), showing the highest improvement of 31.70% as compared to the RT value.展开更多
In the field of satellite imagery, remote sensing image captioning(RSIC) is a hot topic with the challenge of overfitting and difficulty of image and text alignment. To address these issues, this paper proposes a visi...In the field of satellite imagery, remote sensing image captioning(RSIC) is a hot topic with the challenge of overfitting and difficulty of image and text alignment. To address these issues, this paper proposes a vision-language aligning paradigm for RSIC to jointly represent vision and language. First, a new RSIC dataset DIOR-Captions is built for augmenting object detection in optical remote(DIOR) sensing images dataset with manually annotated Chinese and English contents. Second, a Vision-Language aligning model with Cross-modal Attention(VLCA) is presented to generate accurate and abundant bilingual descriptions for remote sensing images. Third, a crossmodal learning network is introduced to address the problem of visual-lingual alignment. Notably, VLCA is also applied to end-toend Chinese captions generation by using the pre-training language model of Chinese. The experiments are carried out with various baselines to validate VLCA on the proposed dataset. The results demonstrate that the proposed algorithm is more descriptive and informative than existing algorithms in producing captions.展开更多
文摘Multi-access interference(MAI)is the main source limiting the capacity and quality of the multiple-input multipleoutput orthogonal frequency division multiplexing(MIMO-OFDM)system which fulfills the demand of high-speed transmission rate and high quality of service for future underwater acoustic(UWA)communication.Therefore,multi-user detection(MUD)is needed at the receiver of the MIMO-OFDM system to suppress the effect of MAI.In this research,MUD is achieved by using a criterion based adaptive recursive successive interference cancellation(RSIC)scheme at the receiver of a MIMO-OFDM system whose transceiver model in underwater communication is implemented by using the Bellhop simulation system.The proposed scheme estimates and eliminates the MAI through user signal detection and subtraction from received signals at the receiver of the MIMO-OFDM system in underwater environment.The bit error rate(BER)performance of the proposed scheme is analyzed by using weight filtering and weight selection criteria.By Matlab simulation,it is shown that the BER performance of the proposed scheme outperforms the conventional matched filter(MF)detector,the adaptive successive interference cancellation(SIC)scheme,and the adaptive RSIC scheme in the UWA network.
基金supported by the National Natural Science Foundation of China (Grant Nos. 51372078 and 51302076)Natural Science Foundation of Hunan Province of China (Grant No. 12JJ4054)+2 种基金Natural Science Foundation of Hunan Province (Grant No. 2018JJ4011)Jiangsu Province Innovative Talent Plan 2016, ChinaYancheng City 515 Talent Plan, China
文摘In the present paper,MoSi2(Cr5Si3)–RSiC composites were prepared via a combination of precursor impregnation pyrolysis(PIP) and MoSi2-Si-Cr alloy active melt infiltration(AAMI) process. Composition, microstructure, mechanical retention characteristics, and oxidation behaviors of the composites at elevated temperature were studied. X-ray diffraction(XRD) pattern confirms that the composites mainly compose of 6 H–SiC, hexagonal MoSi2, and tetragonal Cr5Si3. Scanning electron microscopy(SEM) image reveals that nearly denseMoSi2(Cr5Si3)–RSiC composites exhibiting three-dimensionally(3D) interpenetrated network structure are obtained when infiltrated at 2173 K, and the interface combination of the composites mainly depends on the composition ratio of infiltrated phases. Oxidation weight gain rate of the composites is much lower than that of RSiC matrix, where MoSiCr2 possesses the lowest value of 0.1630 mg×cm-2, about 78% lower than that of RSiC after oxidation at 1773 K for 100 h. Also, it possesses the highest mechanical values of 139.54 MPa(flexural strength σf and RT) and 276.77 GPa(elastic modulus Ef and RT), improvement of 73.73% and 29.77% as compared with that of RSiC, respectively. Mechanical properties of the composites increase first and then decrease with the extension of oxidation time at 1773 K, due to the cooperation effect of surface defect reduction via oxidation reaction and thermal stress relaxation in the composites, crystal growth, and thickness increase of the oxide film. Fracture toughness of MoSiCr2 reaches 2.24 MPa·m1/2(1673 K), showing the highest improvement of 31.70% as compared to the RT value.
基金supported by the National Natural Science Foundation of China (61702528,61806212)。
文摘In the field of satellite imagery, remote sensing image captioning(RSIC) is a hot topic with the challenge of overfitting and difficulty of image and text alignment. To address these issues, this paper proposes a vision-language aligning paradigm for RSIC to jointly represent vision and language. First, a new RSIC dataset DIOR-Captions is built for augmenting object detection in optical remote(DIOR) sensing images dataset with manually annotated Chinese and English contents. Second, a Vision-Language aligning model with Cross-modal Attention(VLCA) is presented to generate accurate and abundant bilingual descriptions for remote sensing images. Third, a crossmodal learning network is introduced to address the problem of visual-lingual alignment. Notably, VLCA is also applied to end-toend Chinese captions generation by using the pre-training language model of Chinese. The experiments are carried out with various baselines to validate VLCA on the proposed dataset. The results demonstrate that the proposed algorithm is more descriptive and informative than existing algorithms in producing captions.
