CN117614849A

CN117614849A - Method and system for sharing multimedia resources based on home network

Info

Publication number: CN117614849A
Application number: CN202410085503.7A
Authority: CN
Inventors: 范绪军; 何建军; 谭金星; 汪贤春
Original assignee: Shenzhen Congxun Intelligent Science And Technology Co ltd
Current assignee: Shenzhen Congxun Intelligent Science And Technology Co ltd
Priority date: 2024-01-22
Filing date: 2024-01-22
Publication date: 2024-02-27

Abstract

The present invention relates to the field of home networks, and in particular, to a method and a system for sharing multimedia resources based on a home network. The method comprises the following steps: acquiring an equipment identification database; carrying out communication mode analysis according to the equipment identification database to obtain an equipment connection relation; performing network demand assessment according to the equipment connection relation to obtain equipment bandwidth demand data; acquiring multimedia resource data; performing model construction on the multimedia resource data by using a convolutional neural network algorithm to generate an automatic multimedia resource identification model; extracting metadata from the multimedia resource to obtain metadata information; index association is carried out on the metadata information and the multimedia resource data to obtain a multimedia database and metadata resource index data; the invention constructs a high-efficiency, safe and user-friendly multimedia service environment by optimizing the resource utilization, improving the user experience, enhancing the data security, ensuring the system performance and the like.

Description

Method and system for sharing multimedia resources based on home network

Technical Field

The present invention relates to the field of home networks, and in particular, to a method and a system for sharing multimedia resources based on a home network.

Background

With the popularization of intelligent home equipment, the home network structure becomes more complex and comprises various connection modes and protocols, the method constructs a highly intelligent home network by utilizing network equipment in the home, such as a router, an intelligent sound box, a television and the like, the equipment can be connected with each other in a wireless or wired mode to form a stable and efficient internal communication system, and the sharing of multimedia resources is not separated from effective management and storage. However, the home network now faces bandwidth limitation, especially when multiple devices access and transmit large-capacity multimedia files simultaneously, bandwidth allocation problems may cause problems such as video clip, audio interruption, and the like, which affect user experience, and maintenance of resource update is troublesome and laborious, and in recent years, bandwidth allocation methods have been used to solve these problems. The intelligent bandwidth management tool can dynamically allocate bandwidth according to the requirements of equipment and network conditions, and is beneficial to preferentially processing key tasks such as video streaming so as to improve user experience. However, bandwidth allocation is generally required when devices are used, and it is impossible to allocate a specific bandwidth capacity to various devices, applications or services before network use, so that network congestion cannot be avoided more.

Disclosure of Invention

Based on this, it is necessary to provide a method and a system for sharing multimedia resources based on a home network, so as to solve at least one of the above technical problems.

In order to achieve the above object, the present invention provides a multimedia resource sharing method based on a home network, comprising the steps of:

step S1: acquiring an equipment identification database; carrying out communication mode analysis according to the equipment identification database to obtain an equipment connection relation; performing network demand assessment according to the equipment connection relation to obtain equipment bandwidth demand data;

step S2: acquiring multimedia resource data; performing model construction on the multimedia resource data by using a convolutional neural network algorithm to generate an automatic multimedia resource identification model; extracting metadata from the multimedia resource to obtain metadata information; index association is carried out on the metadata information and the multimedia resource data to obtain a multimedia database and metadata resource index data;

step S3: data mining is carried out on the multimedia database and the metadata resource index to obtain viewing habit data; performing model construction on the viewing habit data to generate a user preference model; user interest matching is carried out on the database and the resource index by utilizing the user preference model, so that user preference matching data are obtained; performing resource integration on the user preference matching data to obtain intelligent recommendation decision data;

Step S4: based on the equipment bandwidth demand data and the intelligent recommendation decision data, carrying out bandwidth dynamic allocation on the network bandwidth to obtain recombined network bandwidth decision data; performing bandwidth demand association analysis on the device connection relation and the device bandwidth demand data according to the recombined network bandwidth decision data to obtain available cache node data; carrying out buffer priority lifting on available buffer node data to generate preferable buffer node data;

step S5: performing detailed log monitoring on the optimized cache node data to obtain detailed cache log data; carrying out data searchable encryption on the detailed cache log data to obtain encrypted index item data; constructing an index structure of the encrypted index item data to generate searchable encrypted cache node data;

step S6: searching resource data for the searchable encrypted cache node data to obtain an actual cache data set; extracting the resource data characteristics of the detailed cache log data to obtain a record cache data set; performing difference comparison on the actual cache data set and the recorded cache data set to obtain data difference result data; and executing data restoration on the actual cache data set according to the data difference result data.

The method comprises the steps of collecting equipment information connected to a network, including equipment types, MAC addresses, IP addresses and the like, establishing an equipment identification database, recording equipment information, analyzing communication modes among the equipment, knowing interaction relations among the equipment, establishing an equipment connection relation diagram, indicating connection and communication paths among the equipment, evaluating bandwidth requirements of different equipment in the network based on the equipment connection relation to obtain equipment bandwidth requirement data, including bandwidth requirements of each equipment on the network, collecting various multimedia resources, training the multimedia resources by using a Convolutional Neural Network (CNN), constructing an automatic multimedia resource identification model, extracting metadata information from the multimedia resources, establishing index association between the metadata information and the multimedia resource data, forming a multimedia database and metadata resource index data, carrying out data mining on the multimedia database and the metadata resource index, obtaining user watching data, constructing a user preference model by using the watching habit data, matching the database and the resource index by using the user preference model, obtaining user preference habit matching data, integrating the user preference matching data, and forming intelligent decision data. Dynamically allocating network bandwidth based on equipment bandwidth demand data and intelligent recommendation decision data to obtain reorganized network bandwidth decision data, analyzing equipment connection relation and equipment bandwidth demand data according to the network decision data to obtain bandwidth demand association analysis results, obtaining available cache node data, indicating nodes capable of carrying out data caching, carrying out cache priority lifting on the available cache node data to form preferred cache node data, carrying out detailed log monitoring on the preferred cache node data, recording cache activities to obtain detailed cache log data, carrying out data encryption on the detailed cache log data to form encryption index item data, constructing an index structure by utilizing the encryption index item data to form searchable encryption cache node data, carrying out resource data searching by utilizing the searchable encryption cache node data to obtain an actual cache data set, carrying out resource data feature extraction on the detailed cache log data to form a record cache data set, carrying out difference comparison on the actual cache data set and the record cache data set to obtain data difference result data, carrying out data restoration on the actual cache data set according to the data difference result data, and ensuring consistency of the cache data.

The invention has the advantages that the communication mode and the network bandwidth requirement between the devices can be better known by analyzing the connection relation of the devices and evaluating the network requirement, so that a more accurate decision is made for network resource allocation, a multimedia database and a metadata resource index can be established by automatically identifying and extracting the multimedia resources, the subsequent data mining and user interest matching are facilitated, the multimedia database and the metadata resource index can be matched with the interests of the users by analyzing the watching habits of the users and constructing a user preference model, thereby providing personalized intelligent recommendation service, and the network resource utilization rate can be optimized and the effective utilization and dynamic allocation of network resources (such as bandwidth) are ensured by dynamically allocating the network bandwidth according to the device bandwidth requirement data and the intelligent recommendation decision data. Meanwhile, according to priority promotion and availability analysis of the cache nodes, the access efficiency and user experience of data can be improved, detailed cache log data are encrypted, a searchable encryption index structure is constructed, the safety and privacy of the data can be protected, meanwhile, efficient data search and access are realized, and data restoration can be timely carried out through difference comparison between an actual cache data set and a recorded cache data set, so that the integrity and accuracy of the cache data are ensured. This helps to improve the management efficiency of network resources and reduce the delay of data transmission. The dynamic allocation of network bandwidth to the optimization of the cache node and the personalized recommendation of the user are helpful for the system to achieve a better state in terms of resource utilization, response speed and user experience, and the integrity and consistency of the cache data are ensured through the data difference result data and the data restoration steps. Therefore, the invention constructs a high-efficiency, safe and user-friendly multimedia service environment by optimizing the resource utilization, improving the user experience, enhancing the data security, ensuring the system performance and the like.

Preferably, step S1 includes:

step S11: acquiring an equipment identification database;

step S12: drawing a topological structure according to the equipment identification database to obtain an equipment topological structure diagram;

step S13: carrying out communication mode analysis on the equipment topology structure diagram to obtain an equipment connection relation; analyzing the network flow of the equipment connection relation to obtain the data transmission quantity of the equipment;

step S14: and carrying out demand assessment on the data transmission quantity of the equipment to obtain the bandwidth demand data of the equipment.

Preferably, step S2 includes:

step S21: acquiring multimedia resource data;

step S22: constructing a model of the multimedia resource by using a convolutional neural network algorithm to generate an automatic recognition model of the multimedia resource;

step S23: classifying the multimedia resources by utilizing the automatic multimedia resource identification model to obtain grouped multimedia resources;

step S24: extracting metadata from the grouped multimedia resources to obtain metadata information; and carrying out index association on the metadata information and the multimedia resource data to obtain a multimedia database and metadata resource index data.

Preferably, step S3 includes:

step S31: performing user habit data mining on the multimedia database and the metadata resource index data to obtain viewing habit data;

Step S32: data analysis is carried out on the viewing habit data to obtain user preference data;

step S33: performing model construction according to the user preference data to generate a user preference model;

step S34: user interest matching is carried out on the multimedia database and the metadata resource index data by using a user preference model, so that user preference matching data are obtained;

step S35: and carrying out resource integration on the user preference matching data to obtain an intelligent recommendation decision.

Preferably, step S4 includes:

step S41: based on the equipment bandwidth demand data and the intelligent recommendation decision data, dynamically distributing the network bandwidth by utilizing a network bandwidth dynamic adjustment algorithm to obtain a reorganized network bandwidth decision;

step S42: performing bandwidth demand association analysis on the equipment topological structure diagram and the equipment bandwidth demand data according to the recombination network bandwidth decision to obtain available cache node data;

step S43: screening the maximum storage capacity of the available cache node data to obtain the cache node data with the maximum storage capacity;

step S44: and carrying out buffer priority lifting on the buffer node data with the maximum storage capacity to generate preferable buffer node data.

Preferably, the dynamic adjustment algorithm of the network bandwidth in step S42 is as follows

；

Wherein,for bandwidth decision function, ++>For the end time of the integration, +.>At the point in time when the network bandwidth is decided,to +.>Device bandwidth demand at time,/->To +.>Intelligent recommendation decision of moment->For the weight for adjusting the bandwidth requirements of the device, +.>For adjusting influence weights of intelligent recommendations, +.>For the decay rate coefficient for controlling the integral term, +.>A temporary variable representing the integral.

Preferably, step S5 includes:

step S51: performing detailed log monitoring on the preferred cache node to obtain detailed cache log data; carrying out statistical data analysis on the detailed cache log data to obtain the entry cache data;

step S52: carrying out data segmentation on the incoming cache data to obtain cache data blocks; performing data labeling on the cache data blocks to obtain labeled data blocks; carrying out hash calculation on the marked data blocks to obtain data hash values;

step S53: data encryption is carried out on the marked data block and the data hash value, and encrypted index item data are obtained;

step S54: and constructing an index structure of the encrypted index item data based on the node searchable encryption algorithm to generate searchable encrypted cache node data.

Preferably, the node searchable encryption algorithm in step S54 is as follows:

；

Wherein,indicating that the node can search the encrypted cache node is +.>Position and->Status value of time>Indicating that the encryption index item data is +.>Characteristic value of the place>For representing the trend data of the encrypted index item data, < +.>Indicating that the encryption algorithm is +.>Specific performance value at->To describe the encryption algorithm at->Response value of time of day->Indicating that the encryption algorithm is +.>Specific factors of location,/->For encrypting the spatial position of the index item data, +.>Representing node searchable encryption buffer node's state change coefficient over time, +.>Spatial variation coefficient representing index item data, +.>And the evolution coefficient of the variation of the performance of the encryption algorithm with time is represented.

Preferably, step S6 includes:

step S61: searching resource data for the searchable encrypted cache node data to obtain an actual cache data set;

step S62: extracting the resource data characteristics of the detailed cache log data to obtain a record cache data set;

step S63: performing data attribute difference comparison on the actual cache data set and the recorded cache data set to obtain a data difference result;

step S64: carrying out specific difference analysis on the data difference result to obtain specific difference data;

step S65: and carrying out data restoration on the actual cache data set according to the specific difference data.

The invention also provides a multimedia resource sharing system based on the home network, which is used for executing the multimedia resource sharing method based on the home network, and comprises the following steps:

the bandwidth acquisition module is used for acquiring the equipment identification database; carrying out communication mode analysis according to the equipment identification database to obtain an equipment connection relation; performing network demand assessment according to the equipment connection relation to obtain equipment bandwidth demand data;

the database acquisition module is used for acquiring the multimedia resource data; performing model construction on the multimedia resource data by using a convolutional neural network algorithm to generate an automatic multimedia resource identification model; extracting metadata from the multimedia resource to obtain metadata information; index association is carried out on the metadata information and the multimedia resource data to obtain a multimedia database and metadata resource index data;

the preference prediction module is used for carrying out data mining on the multimedia database and the metadata resource index to obtain viewing habit data; performing model construction on the viewing habit data to generate a user preference model; user interest matching is carried out on the database and the resource index by utilizing the user preference model, so that user preference matching data are obtained; performing resource integration on the user preference matching data to obtain intelligent recommendation decision data;

The node selection module is used for dynamically distributing the bandwidth of the network based on the equipment bandwidth demand data and the intelligent recommendation decision data to obtain recombined network bandwidth decision data; performing bandwidth demand association analysis on the device connection relation and the device bandwidth demand data according to the recombined network bandwidth decision data to obtain available cache node data; carrying out buffer priority lifting on available buffer node data to generate preferable buffer node data;

the node encryption module is used for carrying out detailed log monitoring on the optimized cache node data to obtain detailed cache log data; carrying out data searchable encryption on the detailed cache log data to obtain encrypted index item data; constructing an index structure of the encrypted index item data to generate searchable encrypted cache node data;

the data restoration module is used for searching resource data of the searchable encrypted cache node data to obtain an actual cache data set; extracting the resource data characteristics of the detailed cache log data to obtain a record cache data set; performing difference comparison on the actual cache data set and the recorded cache data set to obtain data difference result data; and executing data restoration on the actual cache data set according to the data difference result data.

Drawings

Fig. 1 is a schematic flow chart of steps of a method for sharing multimedia resources based on a home network;

FIG. 2 is a flowchart illustrating the detailed implementation of step S4 in FIG. 1;

FIG. 3 is a flowchart illustrating the detailed implementation of step S5 in FIG. 1;

FIG. 4 is a flowchart illustrating the detailed implementation of step S6 in FIG. 1;

the achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.

Detailed Description

The following is a clear and complete description of the technical method of the present patent in conjunction with the accompanying drawings, and it is evident that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, are intended to fall within the scope of the present invention.

Furthermore, the drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus a repetitive description thereof will be omitted. Some of the block diagrams shown in the figures are functional entities and do not necessarily correspond to physically or logically separate entities. The functional entities may be implemented in software or in one or more hardware modules or integrated circuits or in different networks and/or processor methods and/or microcontroller methods.

It will be understood that, although the terms "first," "second," etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another element. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of example embodiments. The term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.

To achieve the above objective, referring to fig. 1 to 4, a method for sharing multimedia resources based on a home network includes the following steps:

Step S3: performing data mining on the multimedia database and the metadata resource index data to obtain viewing habit data; performing model construction on the viewing habit data to generate a user preference model; user interest matching is carried out on the multimedia database and the metadata resource index data by using a user preference model, so that user preference matching data are obtained; performing resource integration on the user preference matching data to obtain intelligent recommendation decision data;

In the embodiment of the present invention, as described with reference to fig. 1, a flow chart of steps of a home network-based multimedia resource sharing method of the present invention is provided, and in this example, the home network-based multimedia resource sharing method includes the following steps:

in the embodiment of the invention, the identification information of various devices is collected, including the type, the model and the network interface of the devices, the communication mode of each device in the network is analyzed, including the communication protocol, the data transmission mode and the communication frequency, the connection relation between the devices is obtained by analyzing the communication mode and the network topology structure between the devices, including the direct connection relation and the indirect connection relation, the requirement condition of the network for the different devices is evaluated by combining the communication mode and the connection relation of the devices, and the bandwidth requirement data of each device including the bandwidth size and the use mode is obtained according to the network requirement evaluation result.

In the embodiment of the invention, the multimedia resource data can be obtained, which can comprise pictures, audios, videos and the like, the multimedia resource data is preprocessed, which comprises image size standardization, audio format conversion, video frame extraction and the like, a Convolutional Neural Network (CNN) algorithm, such as a common convolutional layer, a pooling layer, a full-connection layer and the like, is used for constructing an automatic multimedia resource identification model, model training is carried out on training data, marked data is used for carrying out supervised learning, a verification set is used for evaluating the model, super parameters are adjusted to optimize model performance, overfitting is prevented, the model can be ensured to be generalized to new multimedia resource data, metadata extraction is carried out on the processed multimedia resource data, such as resolution of images, shooting time, audio time, video frame rate and the like, the extracted metadata information is related to the original multimedia resource data, an index relation is established, an appropriate database system is selected, and the related multimedia resource data and metadata index data are stored in a database.

In the embodiment of the invention, the data record of user watching, listening or browsing is extracted from a multimedia database, the abnormal value and the missing value in the data are processed, the data are sorted according to the dimensionalities of users, resources, time and the like, the characteristics such as watching frequency, resource type preference, watching time period, duration and the like are extracted from watching habit data, a user preference model is constructed by using a machine learning or deep learning algorithm, such as filtering, collaborative filtering, deep neural network and the like based on content, the model is evaluated by using the technologies such as cross verification and the like, model parameters are adjusted to improve accuracy and generalization capability, the historical preference of the users is applied to the index data of the multimedia database and metadata resources, the resources related to the interests of the users are identified, the multimedia resources in the database are screened and matched by using the user preference model to generate user preference matching data, a recommendation system is constructed by using the matching data, and the content-based recommendation, collaborative filtering, the deep learning model and the like can be adopted, and the intelligent decision data is provided for the users according to the obtained personalized recommendation.

In the embodiment of the invention, intelligent recommendation decision data and device bandwidth demand data are integrated to know the influence of user behaviors on network bandwidth, network bandwidth allocation is adjusted in real time based on device bandwidth demands and intelligent recommendation data by adopting a dynamic allocation algorithm, so that tasks with high priority and user recommendation can obtain enough bandwidth resources, the bandwidth allocation situation of each device is recorded to form recombined network bandwidth decision data, connection relations among devices are analyzed based on the recombined network bandwidth decision data, the communication modes and dependency relations among the devices are known, the device bandwidth demand data and the connection relations are combined, the bandwidth demand relations among the devices are analyzed, the bandwidth demands of which devices for certain resources are relatively tight are determined, devices which can be used as cache nodes are determined according to the association analysis results to form available cache node data, a cache priority lifting algorithm is designed based on the device bandwidth demand relations and the intelligent recommendation decision data, the priority of resources related to the user interests on the cache nodes is improved, the lifted cache priority is applied to the available cache node data, and final preferred cache node data is generated.

in the embodiment of the invention, a monitoring system is established, the service condition of a preferred cache node is tracked in real time, relevant information such as access frequency, cache hit rate, data size and the like is recorded, monitored detailed data is recorded as a log, the monitored detailed cache log data is encrypted according to equipment ID, resource ID, time stamp, access type (read/write), cache hit condition and the like, data items needing to be indexed, such as equipment ID, resource ID and the like, are determined so as to support subsequent quick search and retrieval, the index item data in the monitoring log is encrypted to form an encrypted index item data set, an index structure, a tree structure, a hash table and the like which are suitable for the encrypted index item are selected so as to improve search efficiency, the selected index structure is used for organizing and constructing the encrypted index item data, quick and efficient data retrieval is ensured, the encrypted index item data is associated with the encrypted cache log data, consistency between the index item and actual data is ensured, and the encrypted cache log data is combined to generate a final searchable encrypted index item data set.

In the embodiment of the invention, encrypted cache data is obtained from a searchable encryption cache node, search operation is performed by using methods such as character string matching, regular expressions and the like, resource data meeting the conditions is formed into an actual cache data set according to search results, detailed cache log data including information such as access time, operation type and the like of the cache data is obtained from a cache system, characteristic information such as hash value, size, latest access time and the like of the cache data is extracted by analyzing the cache log data, the cache data in the detailed cache log data is formed into a record cache data set according to the result of characteristic extraction, difference comparison is performed by using methods such as hash algorithm, comparison file size and the like, detailed information of data difference including newly added data, lost data and the like is obtained according to the result of difference comparison, and data repair is performed by supplementing missing data, deleting redundant data and the like.

Preferably, step S1 comprises the steps of:

step S11: acquiring an equipment identification database;

step S12: drawing a topological structure of the equipment identification database to obtain an equipment topological structure diagram;

The invention accurately records the device information in the network by constructing the device identification database, is beneficial to comprehensively knowing important information such as device types, quantity and positions in the network, draws a device topology structure diagram, is beneficial to intuitively showing the connection relation and layout among devices, is beneficial to identifying problems such as bottleneck and single-point faults in the network, is beneficial to determining the path and the requirement of data transmission by knowing the communication mode among the devices, provides basis for network optimization and safety setting, can determine how much data is exchanged among the devices by analyzing the network flow, identifies flow peaks and high-load areas, is beneficial to identifying potential bottleneck and performance problems, evaluates the data transmission quantity of the devices, is beneficial to determining the actual requirement of the network on bandwidth, so as to carry out reasonable bandwidth planning and distribution, can avoid bandwidth overload by knowing the bandwidth requirement of the devices, improve the network efficiency and the performance, ensure the stable operation of the network, provide comprehensive knowledge on the network environment, is beneficial to better managing the network, identifying potential risks and taking necessary safety measures, and provides data support for decision-making strategy and making decision more beneficial to the basis.

In the embodiment of the invention, equipment identification information is collected, which comprises equipment ID, MAC address, IP address, equipment type, manufacturer and the like, the collected equipment identification information is stored in a database, the accuracy and the integrity of data are ensured, a network discovery tool or protocol (such as SNMP, LLDP and the like) is used for obtaining the connection information of equipment in a network, topology data are generated, a special network topology drawing software or tool is used for creating a device topology structure diagram based on the collected equipment connection information, the communication mode among the equipment is analyzed, the connection relation and the communication path among the equipment are determined, the direct connection, transit equipment and the like are included, a network monitoring tool or flow analysis software is used for monitoring the data transmission quantity among the equipment, the flow mode and the peak period are identified, and the data transmission quantity among the equipment is identified, and the actual requirement of the equipment on the bandwidth is estimated according to the collected network flow data. And (3) considering factors such as frequency, size, priority and the like of data transmission, planning allocation and optimization of network bandwidth according to an evaluation result, and ensuring that the network can meet the communication requirements among devices.

Preferably, step S2 comprises the steps of:

Step S21: acquiring multimedia resource data;

The invention uses the convolutional neural network to carry out model construction on the multimedia resources, can effectively identify the multimedia resources of different types such as images, videos or audios, improves the accuracy and efficiency of identification, can automatically classify the multimedia resources through the automatic identification model, so that a manager can more easily know and browse a large amount of multimedia data, groups the multimedia resources according to the content characteristics of the multimedia resources, is beneficial to more effectively organizing and managing a large amount of multimedia data, improves the efficiency of data retrieval and browsing, carries out metadata extraction on the grouped multimedia resources, can acquire key information about the resources such as authors, creation dates, places and the like, is beneficial to more comprehensively understanding and managing the multimedia resources, carries out index association on the metadata information and the multimedia resource data, and establishes a multimedia database and metadata resource index data. The index system is beneficial to quick retrieval and positioning of multimedia resources, improves the utilization rate of the resources, enables a user to more easily retrieve needed multimedia resources after having a database and an index system, is convenient for sharing and collaboration of the resources, has the automatic identification and classification capability of the multimedia resources, can be applied to data mining and analysis, finds patterns and trends in data, provides more information support for decision making, can construct an intelligent recommendation system based on classification and metadata information of the multimedia resources, and provides personalized multimedia content recommendation for the user. The management efficiency of the multimedia resources is improved, so that the multimedia resources are easier to understand, utilize and share, and a foundation is provided for further data analysis and application.

In the embodiment of the invention, multimedia data is acquired from various channels, such as a public data set on the internet, data uploaded by a user, real-time video captured by a camera, and the like, the acquired data is preprocessed, including denoising, compression, format conversion and the like, so as to ensure the quality and consistency of the data, the data set is divided into a training set, a verification set and a test set for training and evaluating a model, a convolutional neural network architecture is designed, including a convolutional layer, a pooling layer, a full-connection layer and the like, the model is trained by using the training set, model parameters are continuously adjusted through a back propagation algorithm to minimize a loss function, the performance of the model is evaluated by using the verification set, the model is adjusted and optimized as required, new multimedia resources are automatically identified by using the trained convolutional neural network model, the multimedia resources are classified into different categories according to the output result of the model, the resource classification result is formed, and metadata extraction is performed for the multimedia resources of each group. The metadata may include information of an author, time, place, keywords, etc., the extracted metadata information is associated with the original multimedia resource data to form index data, the multimedia resource data and the associated metadata information are stored in a database to form a multimedia database, and a metadata resource index system is established so that the multimedia resource can be quickly retrieved and located through the metadata information.

Preferably, step S3 comprises the steps of:

According to the invention, through mining the behaviors of the user in the multimedia database, such as clicking, viewing history and the like, the personalized viewing habit of the user is known, including the preferred media types, content topics and the like, the analysis of the viewing habit data can help to construct user portraits, reveal the behavior modes, high-frequency viewing time periods, favorites and the like of the user, a machine learning or deep learning algorithm is utilized to construct a user preference model, the model can predict the favorites of the user on specific multimedia contents, the user preference model is utilized to apply the model to the multimedia database and metadata resource index data, user interest matching is realized, multimedia resources which are most suitable for the user taste are screened out, the opportunity of the user for finding new contents is improved through intelligent recommendation, the viscosity of the user to a platform is enhanced, the user satisfaction degree and experience are improved, the user is encouraged to consume more multimedia contents through personalized recommendation, the activity of the platform is improved, the user interest is satisfied, the more accurate target audience is provided for advertisement delivery through meeting the user interest, the actual behavior feedback of the user can be used for continuously optimizing the recommendation algorithm, and the recommendation accuracy and the user satisfaction degree are improved. The series of steps are beneficial to constructing an intelligent multimedia content recommendation system, so that the personalized requirements of users are better met, and the competitiveness and user experience of a platform are improved.

In the embodiment of the invention, user behavior data is collected from a multimedia database and metadata resource index data, including click records, viewing history and search records, the collected data is cleaned, missing values and abnormal values are processed, necessary timestamp conversion and data format standardization are performed, meaningful features including viewing duration, viewing frequency and favorite media types are extracted from original data, so as to establish feature vectors of user viewing habits, a data mining algorithm such as association rule mining and cluster analysis is used for mining the viewing habits of users, commonalities and differences among different user groups are found, statistical analysis including frequency distribution, time interval distribution and popular content is performed on the mined user viewing habit data, so as to further understand the behavior pattern of the users, analyze the content association in the user viewing history, discover related content which the users may like, provide basis for subsequent recommendation, divide the data sets into training sets and test sets, ensure the reliability of model training, and select proper machine learning or deep learning algorithms such as collaborative filtering and deep neural network according to the nature of the problems. Training a model by using a training set, adjusting model parameters, accurately capturing user preferences, evaluating the model by using a testing set, ensuring generalization capability of the model on new data, applying a trained user preference model to a multimedia database and metadata resource index data, generating personalized interest vectors for each user, calculating similarity between the user interest vectors and multimedia contents, sequencing the similarity by adopting a proper similarity measurement method such as cosine similarity, filtering out contents with larger difference between the user interest vectors, generating personalized interest matching data for the user, integrating the user preference matching data with other related information including heat and freshness of the contents, forming a comprehensively evaluated resource list, formulating recommendation strategies including ranking-based recommendation and content-based recommendation, ensuring that finally recommended multimedia contents conform to the strategy of the user preferences and platforms, collecting feedback information of the user on the recommended contents, and optimizing the model and improving the recommendation accuracy.

Preferably, step S4 comprises the steps of:

According to the invention, through intelligent recommendation decision and equipment bandwidth demand data, the system can dynamically adjust network bandwidth allocation so as to meet real-time bandwidth demands. The method is beneficial to improving network performance, ensuring that a user obtains good experience when watching multimedia content, the system can optimize a connection mode among devices through the correlation analysis of a device topological structure diagram and bandwidth requirements, improve data transmission efficiency, reduce network congestion and delay, select a node with larger storage capacity through screening available cache node data, improve the media content storage capacity of the system, and improve the cache priority of the cache node data with the largest storage capacity, the system can better improve the cache hit rate of related content according to the preference and the requirement of the user, thereby reducing loading time and improving user satisfaction, the comprehensive implementation of the steps is beneficial to creating a comprehensive system, and the efficiency and the performance of the whole multimedia content service system are improved through optimization in multiple aspects of network bandwidth, device topological structure, cache nodes and the like, and the system can more effectively utilize limited network and storage resources through intelligently adjusting the network bandwidth and optimizing the cache nodes, so that the resource utilization efficiency of the whole system is improved.

As an example of the present invention, referring to fig. 2, the step S4 in this example includes:

in the embodiment of the invention, according to the equipment bandwidth demand data and intelligent recommendation decision data, a dynamic adjustment scheme of network bandwidth is generated by using a related algorithm and a model, a dynamic adjustment algorithm of network bandwidth is designed and implemented, according to the real-time bandwidth demand data and recommendation decision data, the network bandwidth allocation is dynamically adjusted to adapt to different workloads and user demands, according to the generated network bandwidth decision, a dynamic bandwidth allocation strategy is implemented, the configuration adjustment of network equipment and the optimization of bandwidth channels are involved, a real-time monitoring mechanism is established, the allocation of the network bandwidth is continuously monitored, the bandwidth allocation is adjusted at any time, the network performance is ensured to always meet the requirements, the dynamically adjusted network bandwidth allocation result is arranged into a form of a reconstructed network bandwidth decision so as to facilitate the execution and analysis of subsequent steps, according to the generated reconstructed network bandwidth decision, the corresponding bandwidth adjustment operation is performed, the network is ensured to operate according to the latest decision, the adjusted network performance is evaluated, the performance index is collected and feedback is provided to improve the future bandwidth decision algorithm and model.

in the embodiment of the invention, topology structure information of each device in a network is acquired, including connection relation and network topology diagram among the devices, the connection relation between the device topology diagram and the device bandwidth demand data are analyzed, the bandwidth demand relation among the devices is determined, the actual network path among the devices is considered, the bandwidth demand relation is evaluated more accurately, the physical path and the logic path of the network topology are considered, based on the result of the bandwidth demand association analysis, which devices in the network can serve as cache nodes are identified, so as to meet the requirement of bandwidth adjustment, the capacity of the available cache nodes is calculated according to the cache capacity and the actual demand of the devices, the load required by bandwidth adjustment is ensured, the identified and calculated available cache node data are output, feedback information about the available cache node data is provided in time in the implementation process, so that monitoring and adjustment are performed, the device bandwidth demand association analysis algorithm is optimized continuously according to the actual application situation, the accuracy and efficiency of association are improved, and the system has certain adaptability and flexibility are ensured by considering the dynamic change of the network topology structure and the bandwidth demand.

in the embodiment of the invention, the available cache node data generated in the step S42 is acquired, the available cache node data is traversed, the storage capacity of each node is compared, a storage capacity threshold is set to determine which nodes have enough storage capacity to meet the requirement of bandwidth adjustment, only the node with the storage capacity larger than the set threshold is reserved, the node is identified as the maximum storage capacity cache node, the screened maximum storage capacity cache node data is output for the subsequent step, the specific process of storage capacity screening is recorded, the used threshold and the screened node information are included, information feedback about the maximum storage capacity cache node is provided for monitoring and adjustment, the threshold of the storage capacity is possibly required to be adjusted according to the actual requirements and system performance, so that better balance is achieved, and the storage capacity screening algorithm is continuously optimized to improve the screening efficiency and accuracy.

In the embodiment of the invention, the standard of buffer priority level promotion is determined, which comprises data access frequency, data importance and request response time factors, the storage amount information of each buffer node is determined, the buffer nodes are ordered, the priority level is determined according to storage amount indexes, a buffer node list needing to be subjected to priority level promotion is determined according to an evaluation result, the priority level of the determined buffer nodes is promoted by adjusting weight parameters in a buffer algorithm and increasing the copy number of buffer data, after the priority level of the buffer nodes is promoted, the data is redistributed according to new preferred buffer node lists, and the data is stored on the preferred buffer nodes, so that the storage condition of the preferred buffer nodes is monitored, and the data is ensured to be stored on the preferred buffer nodes according to expectations.

Preferably, the dynamic adjustment algorithm of the network bandwidth in step S41 is as follows:

；

wherein,for bandwidth decision function, ++>For the end time of the integration, +.>At the point in time when the network bandwidth is decided,to +.>Device bandwidth demand at time,/->To +.>Intelligent recommendation decision of moment- >For adjusting device bandwidth requirementsWeights of->For adjusting influence weights of intelligent recommendations, +.>For the decay rate coefficient for controlling the integral term, +.>A temporary variable representing the integral.

The invention constructs a dynamic adjustment algorithm of network bandwidth, and describes a bandwidth decision function in the algorithmWherein the device bandwidth requirements are related to>And intelligent recommendation decision->Wherein the device bandwidth requirement->Is indicated at +.>The actual bandwidth requirement of the device at the moment is +.>To adjust the weight of the demand, i.e. to control the contribution degree of the device demand to the whole bandwidth decision function, intelligent recommendation decision +.>Is indicated at +.>Time intelligent recommendation decision, algorithm using parameter +.>The influence weight of the intelligent recommendation on the bandwidth decision is adjusted, and the trust degree of the system on the intelligent decision and the suggestion of the system on the whole network resource allocation are reflected. Parameter->Allowing the system to adjust the importance of the device requirements to play an appropriate role in the overall bandwidth decision, parameter +.>The system is allowed to adjust the trust level of the intelligent recommendation, thereby more flexibly considering the recommendation of the intelligent recommendation in the bandwidth decision. / >Is an internal integral representing some effect in time, calculated over time period 0,/for>]An exponential decay function within->Influence on bandwidth decision, wherein ∈>Is a temporary variable of the integration. This section can be used to take into account the effect of the elapsed time to smooth or adjust the change in bandwidth, and the integral term in the formula allows the system to take into account the effect of the elapsed time, thereby achieving dynamic adaptation to the network bandwidth. This ensures that the allocation of bandwidth is not only based on the current time of day demand, but also takes into account historical trends. The network bandwidth dynamic adjustment algorithm constructed by the invention provides a method for comprehensively considering equipment requirements, intelligent recommendation and historical influence so as to make a network bandwidth decision, so that the distribution of network resources is more intelligent, dynamic and adaptive.

Preferably, step S5 comprises the steps of:

The invention monitors the detailed log of the preferred cache node and obtains the data entering the cache through statistical data analysis. The method is beneficial to a system administrator to deeply understand the service condition of the cache node, identify hot spot data and frequently accessed content, and obtain a cache data block and a marked data block by dividing and marking the data entering the cache. This careful data processing enables the system to manage and store data more efficiently, while also providing more information for subsequent operations, helping to uniquely identify and verify data blocks through hash computation. By calculating the hash value of the data block, the integrity of the data can be ensured, the data is prevented from being tampered or damaged, and the marked data block and the data hash value are encrypted to obtain the encrypted index item data. The method enhances the security and privacy protection of the data, builds an index structure based on a searchable encryption algorithm, further improves the security of the data, allows efficient search operation to be performed in an encryption state, and generates searchable encrypted cache node data, so that the system can still perform data retrieval and access in an efficient manner while maintaining the security of the data, and combines the steps of monitoring, analysis, encryption, indexing and the like, and the management efficiency, the data security and the privacy protection level of the cache node are improved through careful data processing and security mechanisms. This is particularly important for systems that face large amounts of sensitive data or for scenarios where it is desirable to ensure data integrity and confidentiality.

As an example of the present invention, referring to fig. 3, the step S5 in this example includes:

in the embodiment of the invention, detailed log monitoring is arranged on a preferred cache node, including enabling a detailed log recording function of a cache system, ensuring that recorded information contains records of key events such as data access, read-write operation and the like, selecting to store the log data on a special log server by using a special log collecting tool or implementing a customized log recording mechanism in the system, also using a distributed storage system, ensuring the integrity and usability of the data, formatting the collected log data for subsequent analysis, including converting the log data into a structured format for query and analysis, analyzing the log data by using a statistical tool or writing a customized script, wherein the analyzed data comprises data access frequency, most commonly accessed data blocks and access modes, extracting the key data entering the cache according to the analysis result, including identifying the data blocks with high access frequency and identifying hot spot data, recording the extracted data entering the cache for later steps, and storing the data in a database or generating specific data files.

in the embodiment of the invention, data entering a cache is subjected to data segmentation according to a preset rule or algorithm, the data is cut into smaller data blocks, each cache data block is segmented based on the principles of data size, fixed-size blocks, specific marking symbols or data structures and the like, each cache data block is marked, the type (such as text, images and videos) of the data block is marked, a time stamp is added to record the creation or modification time of the data block, the source or related information of the data block is marked, custom tags or metadata are added, hash calculation is performed on the marked data block, hash values of the data block are generated, and a hash function maps the content of the data block to a unique identifier with fixed length.

in the embodiment of the invention, a proper encryption algorithm is selected to ensure that the security and performance of the index item data meet the requirements of the system, a symmetric encryption algorithm (such as AES) or an asymmetric encryption algorithm (such as RSA) is generally selected to ensure that a secure key management system is provided, a key required for encryption is generated, distributed, updated and stored, the selected encryption algorithm and the key are used for encrypting the marked data block and the data hash value, the marked data block and the data hash value are realized by calling a corresponding encryption library or tool, the specific realization mode depends on the selected encryption algorithm and programming language, the encrypted marked data block and the encrypted data hash value are combined into encrypted index item data, the encrypted marked data block and the encrypted data hash value and other information are contained, the generated encrypted index item data are stored in a secure position so as to facilitate subsequent retrieval and use, secure communication protocols and encryption channels are ensured to protect the data from being listened to and tampered, and only authorized users or systems can access the index item data by using access tokens, identity verification mechanisms and the like.

In the embodiment of the invention, a node searchable encryption algorithm is utilized to construct a proper index structure to support efficient search operation, including using a tree structure (such as a tree or a tree structure) or other data structures to quickly locate encrypted index item data, further encrypting the encrypted index item data to ensure the security thereof, including using an additional level of encryption provided by the searchable encryption algorithm to prevent information leakage, combining the constructed index structure with the encrypted index item data to generate searchable encrypted cache node data, the cache nodes containing processed index information to support subsequent search operations, storing the generated searchable encrypted cache node data in a proper position so that a system can quickly retrieve and parse the data, ensuring that the system uses a safe communication protocol and an encryption channel to protect the data from eavesdropping and tampering, realizing a support mechanism for the search operation, ensuring that the system can execute effective search on an encrypted data set, performing necessary performance optimization, and ensuring that the search operation can be completed in a reasonable time while keeping the data security.

Preferably, the node searchable encryption algorithm in step S4 is as follows:

；

The invention constructs a node searchable encryption algorithm, whereinIndicating that the node can search the encrypted cache node is +.>Position and->A state value of time, which is a main output of the whole algorithm and reflects the comprehensive influence of the encryption index item data and the performance of the encryption algorithm>For the encryption index item data influence coefficient, this part represents that the encryption index item data is at spatial position +.>Characteristic value>And an exponential function- >And the whole spatial position range (from 0 to +.>) Integration is performed, which describes the trend of the encrypted index item data in space.For the performance impact value of the encryption algorithm, this part represents the time evolution coefficient of the encryption algorithm +.>Performance value>Function of describing response->And a specific factor->And the whole time range (from 0 to +.>) Integration is performed, which in part describes the evolution of the performance of the encryption algorithm over time. The principle of the formula is that the space change trend of the encryption index item data and the time evolution of the encryption algorithm performance are comprehensively considered to calculate the state value of the searchable encryption cache node of the node, wherein the encryption index item data and the encryption algorithm performance are respectively integrated through two integral items and are subjected to integral operation in the whole space and time range, and the formula allows the comprehensive influence of the encryption index item data and the encryption algorithm performance to be dividedBy analyzing, comprehensive performance evaluation is provided for the system, the algorithm can better capture the space-time correlation of the encryption system by considering the change of the space position and time, so that the node state value is closer to the actual situation, and the introduction of parameters (such as characteristic values, performance values, response values, factors and the like) in the formula enables the algorithm to have certain flexibility, and can be adjusted and optimized according to specific application scenes.

Preferably, step S6 comprises the steps of:

According to the invention, the resource data searching is carried out on the searchable encrypted cache node data, so that the accuracy of an actual cache data set can be ensured, the consistency of the data is improved, the specific operation of caching can be recorded by extracting the resource data characteristics of the detailed cache log data, including reading, writing and the like, the detailed operation log is provided for the subsequent data comparison and analysis, and the data attribute difference comparison is carried out on the actual cache data set and the recorded cache data set, thereby being beneficial to detecting the abnormal conditions of the data, such as data loss and damage. The data difference result is subjected to specific difference analysis, so that the problem can be accurately positioned, the situation that the data are changed is known, the abnormal situation is better processed, the data are restored to the actual cache data set according to the specific difference data, the affected data can be restored, the stability and reliability of the system are ensured, the data privacy can be protected by searching encryption, the confidentiality and the integrity of the data can be maintained by monitoring the data difference and timely restoring, the potential problem can be prevented by regularly executing the steps, the system is ensured to be in a good health state, and the risk of system faults caused by data abnormality is reduced.

As an example of the present invention, referring to fig. 4, the step S6 in this example includes:

in the embodiment of the invention, an encryption index of a searchable encryption caching node is acquired by using a proper key or a proper certificate, which relates to an authentication and authorization process, so that only an authorized user or a system can access the encryption index, the encryption index is decrypted by using a corresponding decryption algorithm to obtain original index data, a corresponding search request is constructed according to the characteristics or the identification of resource data to be searched, a search statement or a search condition is constructed, the constructed search request is sent to the searchable encryption caching node according to the design and the realization of the system, the communication protocol and the interface of the caching node are ensured to effectively send the search request and receive the corresponding result, the search result is received from the caching node, the result is decrypted by using the corresponding decryption algorithm, and the decrypted search result is further processed to obtain a final actual caching data set.

In the embodiment of the invention, detailed log data is acquired from a cache system, wherein the log data comprises detailed information about cache operation, such as data reading, writing and updating operation, the cache log data is analyzed to extract useful information, the log data is decomposed into a structure which is easy to process by using a corresponding analyzer or regular expression, and the characteristics of resource data to be extracted are determined according to the system requirements and design, and the characteristics of the resource data comprise the key, the numerical value, the time stamp and the access frequency information of the cache. An algorithm is formulated to extract required resource data features from the parsed log data, specific fields are extracted from the log entries or relevant statistical information is calculated, all the log entries are processed iteratively, a designed feature extraction algorithm is applied to each log data to obtain the features of the resource data, the extracted resource data features are combined to construct a record cache data set, and the constructed record cache data set is stored for subsequent query, analysis or other operations.

In the embodiment of the invention, an actual cache data set and a record cache data set are acquired from a storage medium, a database query, file reading or other modes are involved, corresponding data sources are selected according to actual conditions, data attributes to be compared are determined, including attributes such as cache keys, numerical values, time stamps and the like, format standardization is carried out on the compared data attributes according to system design and requirements to ensure that compared data have consistent data types and formats, algorithms are formulated to compare differences between the two data sets, a hash function algorithm is utilized to quickly detect the differences, corresponding records are marked for the found differences, including specific positions, types and other related information of the recorded differences, a data difference report is generated according to the comparison results, the data difference report can be a data structure containing difference records, can also be in the form of reports, log files and the like, and the generated difference results are stored in proper positions.

in the embodiment of the present invention, the data to be analyzed, including the record marked as the difference, its attribute and other related information, is loaded from the data difference result generated in step S63, and the data difference result is traversed to analyze the difference record one by one. The parsing process includes extracting information of key attributes, difference types, time stamps and the like of the records. Classifying the difference records, classifying the difference records into different categories according to the properties of the differences, including data missing, numerical mismatch and inconsistent time stamps, associating the difference records with related context information, including sources of records, corresponding business operations and system state information, detecting abnormal values of numerical data, identifying abnormal constant values causing the differences, formulating specific solving strategies for each difference type, including data restoration, system adjustment and alarm notification, and generating specific difference data reports, including statistical information of the differences, proposal of solving strategies and content of operation proposal according to analysis results.

In the embodiment of the invention, the priority of repair is confirmed according to the classification and service requirements of specific difference data, original buffer data is selected to be backed up before data repair is carried out, corresponding repair strategies are formulated for each specific difference condition, the buffer configuration is adjusted, the actual buffer data set is repaired according to the formulated repair strategies, database inquiry, data conversion and buffer refreshing operations are involved, the specific method depends on the realization and architecture of a buffer system, after the data repair is completed, the repair effect is verified, the difference analysis is rerun, the consistency of the repaired data and source data is ensured, the difference problem is solved, and the data repair process including repair time, repair data range and specific measure information is recorded.

In the present specification, there is provided a home network-based multimedia resource sharing system for performing the home network-based multimedia resource sharing method as described above, comprising:

The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

The foregoing is only a specific embodiment of the invention to enable those skilled in the art to understand or practice the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. The multimedia resource sharing method based on the home network is characterized by comprising the following steps:

2. The home network-based multimedia resource sharing method according to claim 1, wherein the step S1 comprises the steps of:

step S11: acquiring an equipment identification database;

3. The home network-based multimedia resource sharing method according to claim 1, wherein the step S2 comprises the steps of:

step S21: acquiring multimedia resource data;

4. The home network-based multimedia resource sharing method according to claim 1, wherein the step S3 comprises the steps of:

5. The home network-based multimedia resource sharing method as claimed in claim 1, wherein the step S4 comprises the steps of:

6. The home network-based multimedia resource sharing method of claim 5, wherein the network bandwidth dynamic adjustment algorithm in step S41 is as follows:

；

Wherein,for bandwidth decision function, ++>For the end time of the integration, +.>Time point for network bandwidth decision, +.>To +.>Device bandwidth demand at time,/->To +.>Intelligent recommendation decision of moment->For the weight for adjusting the bandwidth requirements of the device, +.>For adjusting influence weights of intelligent recommendations, +.>For the decay rate coefficient for controlling the integral term, +.>A temporary variable representing the integral.

7. The home network-based multimedia resource sharing method according to claim 1, wherein the step S5 comprises the steps of:

8. The home network-based multimedia resource sharing method as claimed in claim 7, wherein the node searchable encryption algorithm in step S54 is as follows:

；

9. The home network-based multimedia resource sharing method as claimed in claim 1, wherein the step S6 comprises the steps of:

10. A home network-based multimedia resource sharing system for performing the home network-based multimedia resource sharing method of claim 1, comprising: