Настройки

Укажите год
-

Небесная энциклопедия

Космические корабли и станции, автоматические КА и методы их проектирования, бортовые комплексы управления, системы и средства жизнеобеспечения, особенности технологии производства ракетно-космических систем

Подробнее
-

Мониторинг СМИ

Мониторинг СМИ и социальных сетей. Сканирование интернета, новостных сайтов, специализированных контентных площадок на базе мессенджеров. Гибкие настройки фильтров и первоначальных источников.

Подробнее

Форма поиска

Поддерживает ввод нескольких поисковых фраз (по одной на строку). При поиске обеспечивает поддержку морфологии русского и английского языка
Ведите корректный номера.
Ведите корректный номера.
Ведите корректный номера.
Ведите корректный номера.
Укажите год
Укажите год
Применить Всего найдено 6. Отображено 6.
30-06-2023 дата публикации

Optical fiber nonlinear equalization method based on WideDeep-CNN (Convolutional Neural Network)

Номер: CN116366165A
Автор: LIU XINYU, XIN XIANGJUN, ZHOU SITONG, GAO RAN, CHANG HUAN, BI JIAHAO, WANG YONGJUN, ZHANG QI, GUO DONG, LI ZHIPEI, WANG FU, ZHANG WENQUAN, WU RUIDE, HUANG XIN
Принадлежит:

According to the invention, the method is based on Wideamp; the invention discloses an optical fiber nonlinear equalization method of a Deep-CNN, and belongs to the technical field of optical fiber communication. The implementation method comprises the following steps: constructing a first characteristic sequence and a second characteristic spectrum of each M-QAM signal, and constructing a training data set; the construction is based on the Wideamp; a nonlinear equilibrium model of the Deep-CNN model is adopted, and the first feature sequence is adopted as a Wideamp; an input feature sequence of a Wide network sub-model in the Deep-CNN model, and the second feature spectrum are used as Wide amp; an input characteristic spectrum of a Deep-CNN sub-model in the Deep-CNN model is obtained; using the training data set to train the Widamp; a Deep-CNN model is used for training; the features of each M-QAM signal to be subjected to nonlinear equalization are input into the trained Wideamp; the ...

Подробнее
Номер записи: 1
28-04-2023 дата публикации

Method and device for determining multimedia object, equipment and storage medium

Номер: CN116028656A
Автор: LI ZHEN, BI JIAHAO, GUO SHANGFENG
Принадлежит:

The invention discloses a method, device and equipment for determining a multimedia object and a storage medium. The method comprises the steps that when it is detected that a multimedia object is watched, associated information of the multimedia object is acquired; determining at least one first to-be-displayed multimedia object associated with the name information based on the name information and preset multimedia object series information; determining at least one second to-be-displayed multimedia object associated with the name information based on the name information and a pre-constructed multimedia object knowledge graph; determining at least one target multimedia object based on the first to-be-displayed multimedia object, the second to-be-displayed multimedia object and the user portrait; according to the method, the at least one target multimedia object is displayed, and the related multimedia object is searched and displayed for the user in real time according to the associated ...

Подробнее
Номер записи: 2
18-04-2023 дата публикации

Photon radio frequency signal generation method based on pre-coding and probabilistic shaping joint optimization

Номер: CN115987404A
Автор: GUO DONG, LI XINYING, GAO RAN, XIN XIANGJUN, YAN HENGXIN, BI JIAHAO, PAN XIAOLONG, YU CHAO, ZHANG QI, YAO HAIPENG, DONG ZE, ZHANG WENQUAN, WU RUIDE, HUANG XIN, YAN JINGHAO
Принадлежит:

The invention discloses a method for generating a photon radio frequency signal based on precoding and probability shaping joint optimization, and belongs to the field of communication. The implementation method comprises the following steps of: splitting a binary information sequence into a basic layer sequence and a processing layer sequence; a processing layer sequence is input into a distribution matcher to generate a non-uniform symbol sequence with expected probability distribution, a quadrant bit and an information bit are combined to realize probability shaping, and the information sequence is converted into a PS-QAM signal through QAM modulation; pre-processing the amplitude and phase of the signal by using a pre-coding technology; after the modulation signal and the radio frequency sine wave signal are mixed, a Mach-Zehnder modulator is driven to realize optical carrier suppression modulation on continuous optical waves generated by a distributed feedback laser; optical millimeter ...

Подробнее
Номер записи: 3
18-04-2023 дата публикации

Optical fiber nonlinear equalization method based on WideDeep model

Номер: CN115987394A
Автор: LIU XINYU, XIN XIANGJUN, ZHOU SITONG, GAO RAN, CHANG HUAN, BI JIAHAO, WANG YONGJUN, ZHANG QI, GUO DONG, LI ZHIPEI, WANG FU, ZHANG WENQUAN, WU RUIDE, HUANG XIN
Принадлежит:

The invention discloses a wireless communication method based on a Wideamp; the invention discloses an optical fiber nonlinear equalization method of a Deep model, and belongs to the technical field of optical fiber communication. The implementation method comprises the following steps: constructing a first feature sequence and a second feature sequence of each M-QAM signal, and constructing a training data set; the construction is based on the Wideamp; a nonlinear equilibrium model of the Deep model is adopted, and the first feature sequence is adopted as Wideamp; an input feature sequence and a second feature sequence of a Wide network sub-model in the Deep model are used as Wide amp; an input feature sequence of a Deep network sub-model in the Deep model; using the training data set to train the Widamp; the Deep model is trained; the feature sequence of each M-QAM signal to be subjected to nonlinear equalization is input into the trained Wideamp; the Deep model is used for outputting ...

Подробнее
Номер записи: 4
27-06-2023 дата публикации

Product recommendation method and device, electronic equipment and readable storage medium

Номер: CN116342222A
Автор: LI ZHEN, BI JIAHAO
Принадлежит:

The invention provides a product recommendation method and device, electronic equipment and a readable storage medium. The method comprises the following steps: obtaining a label set of each product in a product recommendation pool, and grouping all products in the product recommendation pool based on the label set of each product to obtain at least one product queue; for each product queue, obtaining product click data of each product in the product queue within a target recommendation duration, and determining corresponding product payment data based on the product click data; determining to-be-recommended products in all the products in the product queue based on the product payment data corresponding to each product, and arranging the to-be-recommended products to obtain a target queue; and recommending all the to-be-recommended products in the target queue based on the sorting of each to-be-recommended product in the target queue. Therefore, the economic benefit in the video consumption ...

Подробнее
Номер записи: 5
23-06-2023 дата публикации

Recommendation method and device based on user preference, equipment and storage medium

Номер: CN116320549A
Автор: LI ZHEN, BI JIAHAO, GUO SHANGFENG
Принадлежит:

The invention relates to the field of content recommendation, in particular to a recommendation method and device based on user preferences, equipment and a storage medium. The method comprises the steps of obtaining preference information of a user, obtaining a film watching label of the user, calculating a real-time evaluation value corresponding to the film watching label according to the film watching label and a preset rule, obtaining right and interest information of the user, and determining a recommendation stream corresponding to the user according to the real-time evaluation value and the right and interest information. According to the method, recommendation is carried out on different users in different recommendation modes, if it is detected that the preference information contains the preference content, the film watching label value is automatically calculated, and the user preference recommendation stream is constructed according to the label value, so that the construction ...

Подробнее
Номер записи: 6