CN111984364B - Artificial intelligence cloud platform towards 5G age - Google Patents
Artificial intelligence cloud platform towards 5G age Download PDFInfo
- Publication number
- CN111984364B CN111984364B CN201910426197.8A CN201910426197A CN111984364B CN 111984364 B CN111984364 B CN 111984364B CN 201910426197 A CN201910426197 A CN 201910426197A CN 111984364 B CN111984364 B CN 111984364B
- Authority
- CN
- China
- Prior art keywords
- distributed
- cloud computing
- computing
- mobile
- server
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
- G06F9/505—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering the load
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
-
- G—PHYSICS
- G08—SIGNALLING
- G08C—TRANSMISSION SYSTEMS FOR MEASURED VALUES, CONTROL OR SIMILAR SIGNALS
- G08C17/00—Arrangements for transmitting signals characterised by the use of a wireless electrical link
- G08C17/02—Arrangements for transmitting signals characterised by the use of a wireless electrical link using a radio link
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/51—Discovery or management thereof, e.g. service location protocol [SLP] or web services
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/4557—Distribution of virtual machine instances; Migration and load balancing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/45595—Network integration; Enabling network access in virtual machine instances
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The invention discloses an artificial intelligent cloud platform oriented to the 5G age, which particularly relates to the fields of artificial intelligence, cloud computing and big data, and comprises a heterogeneous distributed cloud platform, wherein the heterogeneous distributed cloud platform takes a distributed mobile cloud computing collaborative architecture as a basic architecture; the distributed mobile cloud computing collaborative architecture comprises an intelligent terminal, a mobile network, a distributed mobile cloud computing server, a collaborative controller and an end-to-end computing offloading service quality assurance mechanism system. The invention establishes a 5G-oriented distributed mobile cloud computing collaborative architecture and an end-to-end-terminal computing offloading service quality assurance mechanism, which can reduce network interaction signaling overhead of the terminal in computing offloading and CPU occupancy rate and energy consumption of a terminal residence decision process; the end-to-end service quality is guaranteed, a customized and expanded distributed deep learning platform of a deep learning model and an algorithm is supported, and huge load and energy consumption brought by traditional mobile cloud computing to a wide area network can be avoided.
Description
Technical Field
The invention relates to the technical fields of artificial intelligence, cloud computing and big data, in particular to an artificial intelligence cloud platform oriented to the 5G age.
Background
Artificial intelligence (artificia intelligence), english is abbreviated AI. The method is a new technical science for researching and developing theories, methods, technologies and application systems for simulating, extending and expanding the intelligence of people;
the 5G network is a fifth generation mobile communication network, the peak theoretical transmission speed of the network can reach 1Gb per second, which is hundreds of times faster than the transmission speed of the 4G network, and along with the birth of 5G technology, the era of sharing 3D movies, games and ultra-high quality (UHD) programs with intelligent terminals is going to our day;
the universal interconnection era driven by 5G will generate massive data, and the demand for cloud computing will increase accordingly; the 5G edge calculation can better realize sensing, interaction and control between objects through the data processing capacity closer to the application side, and a huge increment space is brought to cloud calculation.
As data tsunami grows, the demand for computing power will increase greatly, and there is no good solution and artificial intelligence cloud platform for explosive information growth and dynamic flexible architecture requirements.
Disclosure of Invention
In order to overcome the defects of the prior art, the embodiment of the invention provides an artificial intelligent cloud platform oriented to the 5G age, which is characterized in that a 5G oriented distributed mobile cloud computing collaborative architecture and an end-to-end computing offloading service quality assurance mechanism system are established by independently developing a 5G oriented high-performance heterogeneous distributed cloud platform and fusing a distributed mobile cloud computing collaborative architecture technology, so that network interaction signaling overhead of a terminal in computing offloading and CPU occupation rate and energy consumption of a terminal residence decision process can be reduced; the distributed deep learning platform supporting the custom expansion of the deep learning model and the algorithm can ensure the end-to-end service quality of mobile cloud computing, can avoid huge load and energy consumption brought by the traditional mobile cloud computing to a wide area network, carries out real-time interaction through a 5G network and a cloud, improves data processing capacity, reduces time delay, breaks through the bottleneck of an AI transmission technology by fusing a 5G technology, and realizes intelligent energization.
In order to achieve the above purpose, the present invention provides the following technical solutions: an artificial intelligence cloud platform oriented to the 5G age comprises a heterogeneous distributed cloud platform, wherein the heterogeneous distributed cloud platform takes a distributed mobile cloud computing collaborative architecture as a basic architecture;
the distributed mobile cloud computing collaborative architecture comprises an intelligent terminal, a mobile network, a distributed mobile cloud computing server, a collaborative controller and an end-to-end computing offloading service quality assurance mechanism system, wherein the distributed mobile cloud computing server is connected with a service base station, and the collaborative controller is connected with the collaborative server;
the intelligent terminal is used as an initiating terminal of mobile cloud computing, periodically uploads state sensing information of the intelligent terminal through a mobile communication network and receives a relevant unloading segmentation decision calculated by a cooperative controller;
the mobile network provides wireless access and transmission for an intelligent terminal which initiates a calculation unloading request;
the distributed mobile cloud computing server is deployed on a small server or a multi-small server cluster at a mobile access network side, the load state and the virtual machine computing capacity perception information are periodically uploaded to a cooperative server, and the cooperative controller decision information is received to reserve virtual machine resources for a computing unloading task;
the cooperative controller is used for collecting the perception information of the intelligent terminal, the mobile network and the distributed mobile cloud computing server, generating a calculation unloading segmentation decision and transmitting the calculation unloading segmentation decision to the intelligent terminal, and transmitting a resource reservation decision to the service base station and the distributed mobile cloud computing server;
the end-to-end-terminal computing unloading service quality assurance mechanism system comprises a distributed cloud computing perception module and a collaborative decision-making module, wherein the distributed cloud computing perception module respectively works at a server level and a virtual machine level, and the collaborative decision-making module comprises a mobile terminal part decision-making information unit, a mobile communication network part decision-making information unit and a distributed cloud computing node decision-making information unit;
the heterogeneous distributed cloud platform comprises a heterogeneous distributed artificial intelligent cloud computing center, a distributed deep learning platform, a deep learning large-scale training system, a heterogeneous super computing platform and a heterogeneous basic algorithm library;
the deep learning large-scale training system is used for multi-machine multi-CPU-FPGA-GPU hybrid distributed deep learning model training, supports a model with trillion-level parameters, and carries out large-scale classification of hundreds of billions;
the heterogeneous super computing platform provides a plurality of computing clusters, central unified storage and lightweight virtualization, and provides continuous computing capability support for researchers;
the heterogeneous basic algorithm library stores various machine learning algorithms and mathematical and image processing algorithms including a deep neural network;
the heterogeneous distributed artificial intelligent cloud computing center is used for realizing artificial intelligent real-time interaction between a 5G network and a cloud;
the distributed deep learning platform is used for supporting the customized extension of a deep learning model and an algorithm and supporting CPU-GPU or GPU-GPU or CPU-FPGA-GPU hybrid distributed operation.
In a preferred embodiment, the cooperative controllers are in the form of instances or virtual machines running in distributed cloud computing servers or other network elements coexisting with operators, in particular service gateways, packet data gateways and policy and resource management modules.
In a preferred embodiment, the intelligent terminal is connected to a mobility management entity through a local gateway, the mobility management entity is connected to a packet data gateway and a policy and resource management device through a service gateway, the packet data gateway and the policy are connected to the resource management device, the packet data gateway and the policy and the resource management device are respectively connected to an operation business terminal and the internet, and the intelligent terminal is connected to a distributed mobile cloud computing server through the local gateway.
In a preferred embodiment, the distributed cloud computing awareness module is used for collecting the load condition of the whole computing node server or the server cluster at the server level, and specifically includes server throughput, server concurrent communication status, server computing resource usage condition, server storage resource occupancy rate, and when the computing node is the server cluster and adopts a virtualization technology to implement a virtualized resource pool of the whole cluster in the cluster, the server level awareness information should fully consider the overall condition of the virtual resource pool.
In a preferred embodiment, the distributed cloud computing perception module is configured to collect virtual machine state information in the entire node at the virtual machine level, where the information includes: the number of virtual machines, the amount of computing and storage resources occupied by each virtual machine, throughput produced by each virtual machine, and status information related to the bearer resources.
In a preferred embodiment, the cooperative controller fully grasps state information of the intelligent terminal, the mobile network and each distributed cloud computing node by collecting perception information of the intelligent terminal, the mobile communication network and the distributed mobile cloud computing server, and the cooperative decision module generates cooperative decisions by comprehensively analyzing the grasped information and respectively transmits the cooperative decisions to the intelligent terminal, the mobile network element and the distributed cloud computing node, and each part executes respective corresponding actions according to the decisions.
In a preferred embodiment, the mobile terminal part decision information unit is configured to determine a calculation subtask division of a corresponding mobile application according to a current battery, energy consumption and calculation resource state of the intelligent terminal, and further plan a local calculation task and offload calculation tasks according to a wireless bandwidth resource state of the intelligent terminal, when the terminal accesses to multiple base stations, the mobile terminal part decision information unit is related to designating a corresponding access base station for each offload calculation subtask, and at this time, the mobile terminal part decision information unit correctly reassembles calculation subtask results returned in an out-of-order manner.
In a preferred embodiment, the mobile communication network part decision information unit is configured to allocate corresponding access points for unloading calculation task data and returning result data to complete receiving and sending according to a wireless bandwidth state between the intelligent terminal and each access base station and a backhaul network congestion state of each base station.
In a preferred embodiment, the distributed cloud computing node decision information unit is configured to assign, according to status information of each distributed cloud computing node, a corresponding distributed cloud computing node to each offload computing subtask to complete a computing load task; the method comprises the steps that each distributed cloud computing node is used for deciding that each distributed cloud computing node bears virtualized resources required by corresponding unloading computing sub-tasks, and a virtual machine is generated to complete the computing tasks; the parallel collaboration is realized for the computing processes among a plurality of distributed cloud computing nodes.
The invention has the technical effects and advantages that:
1. according to the invention, by aiming at the requirements of enhancing mobile broadband, low time delay, high reliability and large connection and low power consumption in the application scene of the AI cloud computing service system in the 5G era, the defects of the existing network cloud computing service platform in technical architecture are improved, the 5G era oriented high-performance heterogeneous distributed cloud platform is independently researched and developed, the distributed mobile cloud computing collaborative architecture technology is fused, a 5G oriented distributed mobile cloud computing collaborative architecture and an end-to-end computing offloading service quality assurance mechanism system are established, and network interaction signaling overhead of a terminal in computing offloading and CPU occupation rate and energy consumption of a terminal residence decision process can be reduced; the end-to-end service quality of mobile cloud computing can be ensured, a customized and expanded distributed deep learning platform of a deep learning model and an algorithm is supported, and huge load and energy consumption brought by the traditional mobile cloud computing to a wide area network can be avoided;
2. in the 5G era, the invention adopts a high-performance heterogeneous distributed cloud platform, combines a deep learning large-scale training system, utilizes a heterogeneous high-performance computing center and a high-performance heterogeneous basic algorithm library to provide a powerful enterprise artificial intelligence innovation service solution for the 5G intelligent era for users, fully utilizes the characteristics of high speed and low delay of 5G, promotes intelligent cooperation of a cloud AI and a terminal AI, carries out real-time interaction through a 5G network and the cloud, improves data processing capacity, reduces delay, breaks through the bottleneck of an AI transmission technology by a fused 5G technology, and realizes intelligent energization;
3. the terminal side AI can quickly respond to the user demand, and quickly display the processed image, video, voice and text information to the user in a low-power consumption and low-cost mode, so that the terminal side AI is suitable for completing AI reasoning tasks; the integration of the 5G technology enables intelligent collaborative innovation integration of the cloud terminal AI and the terminal side AI to be possible, real-time interaction is carried out through the 5G network and the cloud terminal, the data processing capacity is improved, and the time delay is reduced; the cloud AI is used for realizing multi-terminal data aggregation, has more advantages in the aspects of data throughput, processing speed and the like, and is suitable for completing large-scale large-data-volume AI model training tasks.
Drawings
Fig. 1 is a block diagram showing the overall structure of the present invention.
Fig. 2 is a block diagram of a distributed mobile cloud computing architecture according to the present invention.
Fig. 3 is a topology diagram of a distributed mobile cloud computing collaborative architecture of the present invention.
Fig. 4 is a topology diagram of an end-to-end computing offload quality of service assurance mechanism system of the present invention.
Fig. 5 is a schematic structural diagram of a distributed deep learning platform according to the present invention.
FIG. 6 is a schematic diagram of a deep learning large scale training system according to the present invention.
FIG. 7 is a schematic diagram of a heterogeneous computing platform according to the present invention.
FIG. 8 is a schematic diagram of a heterogeneous underlying algorithm library architecture of the present invention.
The reference numerals are: the system comprises a 1 heterogeneous distributed cloud platform, a 11 heterogeneous distributed artificial intelligent cloud computing center, a 12 distributed deep learning platform, a 13 deep learning large-scale training system, a 14 heterogeneous super computing platform, a 15 heterogeneous basic algorithm library, a 2 distributed mobile cloud computing collaborative architecture, a 21 intelligent terminal, a 22 mobile network, a 23 distributed mobile cloud computing server, a 24 collaborative controller, a 25 end-to-end-terminal computing offloading service quality assurance mechanism system and a 3 service base station.
Description of the embodiments
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
According to the artificial intelligence cloud platform facing the 5G age shown in the figures 1-3, the artificial intelligence cloud platform comprises a heterogeneous distributed cloud platform 1, wherein the heterogeneous distributed cloud platform 1 takes a distributed mobile cloud computing cooperative framework 2 as a basic framework;
the mobile cloud computing offloads the computing task of the terminal to the cloud, reduces the energy consumption of the terminal, and provides the mobile application capability with large computing capacity under the condition of limited resources of the terminal, so that the mobile cloud computing becomes an indispensable key technology for constructing future mobile Internet innovative services; however, in the current technology development of mobile cloud computing, for the problems of unstable end-to-end network time delay and bandwidth from a mobile device to a cloud computing center, dynamic performance of main factors influencing the quality of mobile cloud computing service and the like, an effective countermeasure is still lacking, and meanwhile, effective mechanism design is lacking in terms of terminal perception and negotiation decision for computing and unloading, so that the invention provides a distributed mobile cloud computing collaborative architecture 2 by combining the characteristics of 4G and 5G network wireless access network technologies;
the distributed mobile cloud computing collaborative architecture 2 comprises an intelligent terminal 21, a mobile network 22, a distributed mobile cloud computing server 23, a collaborative controller 24 and an end-to-end computing offloading service quality assurance mechanism system 25, wherein the distributed mobile cloud computing server 23 is connected with a service base station 3, and the collaborative controller 24 is connected with the collaborative server;
the intelligent terminal 21 is used as an initiating terminal of mobile cloud computing, periodically uploads state sensing information of the intelligent terminal through a mobile communication network and receives relevant unloading segmentation decisions calculated by the cooperative controller 24;
the mobile network 22 provides wireless access and transmission for the intelligent terminal 21 that initiates the computing offload request;
the distributed mobile cloud computing server 23 is deployed on a mini-server or a multi-mini-server cluster at the mobile access network side, and the load state and the virtual machine computing capacity perception information thereof are periodically uploaded to a cooperative server and received to the cooperative controller 24 to make decision information so as to reserve virtual machine resources for computing and unloading tasks;
the cooperative controller 24 is configured to collect sensing information of the intelligent terminal 21, the mobile network 22, and the distributed mobile cloud computing server 23, generate a calculation offloading segmentation decision, and send the calculation offloading segmentation decision to the intelligent terminal 21, and send a resource reservation decision to the service base station 3 and the distributed mobile cloud computing server 23;
the cooperative controller 24 is in the form of an instance or a virtual machine running in a distributed cloud computing server or other network elements coexisting with an operator, specifically a service gateway, a packet data gateway and a policy and resource management module;
the intelligent terminal 21 is connected with a mobility management entity through a local gateway, the mobility management entity is connected with a packet data gateway and a policy and resource management device through a service gateway, the packet data gateway and the policy are connected with the resource management device, the packet data gateway and the policy are respectively connected with an operation business terminal and the internet, and the intelligent terminal 21 is connected with a distributed mobile cloud computing server 23 through the local gateway;
the implementation mode specifically comprises the following steps: by fusing the technology of the distributed mobile cloud computing collaborative architecture 2, a 5G-oriented distributed mobile cloud computing collaborative architecture 2 and an end-to-end computing offloading service quality assurance mechanism system 25 are established, so that network interaction signaling overhead of a terminal in computing offloading and CPU occupancy rate and energy consumption of a terminal residence decision process can be reduced; the end-to-end service quality of mobile cloud computing can be ensured; the huge load and energy consumption brought by the traditional mobile cloud computing to the WAN can be avoided,
it is anticipated that, under the multi-party push of the user, the smart device manufacturer and the mobile network 22 operator, future distributed mobile cloud computing will become one of the mainstream technologies of the 4G and 5G networks, become the infrastructure of the mobile operator and provide a new service growth point for the mobile operator, and provide a better user experience for the application of the smart terminal 21.
Although the current 5G mobile network 22 technology is still under pre-development, numerous potential technologies that 5G networks can foresee provide effective feasibility support for the distributed mobile cloud computing proposed above, including:
a. ultra dense bee deployment
The idea of deployment of micro base stations of an LTE4G network, such as femto base station (Femtocell) base station, pico base station (Picocell) and the like, is continued for the purpose of improving network coverage quality and network capacity. Under the condition that the deployment of the 5G network micro base stations is denser and the same-frequency interference and wireless resource multiplexing problem is effectively solved, the wireless networking mode can bear more mobile network 22 transmission, and the energy consumption and time delay of wireless data transmission of the intelligent terminal 21 in mobile cloud computing can be effectively reduced due to the fact that the average wireless transmission distance between the access point and the intelligent terminal 21 is shortened.
b. Massive MIMO
As a derivative enhancement of the multiple-input multiple-output (MIMO) technology, the large-scale MIMO (MassiveMIMO) can obtain a substantial improvement of the spatial beamforming gain through a simple linear precoding algorithm by deploying a larger number (tens or hundreds) of antenna units at the transmitting end and/or the receiving end, so as to significantly enhance the communication reliability, link throughput, spectral efficiency and energy efficiency of the point-to-point/multipoint link, thereby being capable of carrying more concurrent mobile cloud computing transmission at a high rate.
c. Millimeter wave backhaul
With the proliferation of the number of base stations, the backhaul network connecting macro base stations, micro base stations and mobile switching nodes becomes a key for ensuring network performance. Currently, millimeter waves (including 71-76, 81-86, 92-95 GHz frequency bands, etc.) are being used as the carrier for 4G backhaul networks, which are also being further enhanced in 5G networks. The microwave backhaul provides guarantee for backhaul bandwidth and time delay of macro base stations and micro base stations in the 5G network, and particularly can effectively solve the problem of delay jitter and congestion of ADSL backhaul links of the micro base stations, so as to support the perceived uploading of the intelligent terminal 21 and the base stations to the cooperative controller 24 in the distributed mobile cloud computing cooperative structure with low delay and high speed, and the decision issuing of the cooperative controller 24 to the terminal and the base stations.
As shown in fig. 4, the computing task of the terminal offloaded by mobile cloud computing is finally required to be completed in the virtual machines in the distributed cloud computing servers distributed in the mobile access network, so that the distributed cloud computing perception including the load of each distributed computing node, the number of the virtual machines which can be borne in each node and the computing capability of each virtual machine is an important basis for selecting the computing offloaded node, and has an important influence on the computing performance of the mobile cloud;
in the foregoing, the end-to-end computing offload qos guarantee mechanism system 25 includes a distributed cloud computing awareness module and a collaborative decision module, where the distributed cloud computing awareness module works at a server level and a virtual machine level, and the collaborative decision module includes a mobile terminal part decision information unit, a mobile communication network part decision information unit, and a distributed cloud computing node decision information unit;
the distributed cloud computing perception module is used for acquiring the load condition of a whole computing node server or a server cluster at a server level, and specifically comprises server throughput, server concurrent communication state, server computing resource use condition, server storage resource occupancy rate and when the computing node is the server cluster and a virtualization technology is adopted in the cluster to realize a virtualization resource pool of the whole cluster, the server level perception information fully considers the whole condition of the virtual resource pool;
the distributed cloud computing perception module is used for collecting virtual machine state information in the whole node at the virtual machine level, and the information comprises: the number of virtual machines, the amount of calculation and storage resources occupied by each virtual machine, throughput produced by each virtual machine in an equivalent way and related state information of bearing resources;
at present, the main stream server products all support the software acquisition and storage of the information, and the distributed cloud computing high-efficiency perception of the cooperative controller is realized through a high-efficiency interface and information sharing mechanism;
the cooperative controller 24 fully grasps the state information of the intelligent terminal 21, the mobile network 22 (including an access network and a core network) and each distributed cloud computing node by collecting the perception information of the intelligent terminal 21, the mobile communication network and the distributed mobile cloud computing server 23, and the cooperative decision module generates a cooperative decision by comprehensively analyzing the grasped information and respectively transmits the cooperative decision to the intelligent terminal 21, the mobile network 22 network element and the distributed cloud computing node, and each part executes respective corresponding actions according to the decision to realize cooperative distributed cloud computing and provide user experience guarantee for terminal application;
the mobile terminal part decision information unit is used for determining the calculation subtask division of corresponding mobile application according to the current battery, energy consumption and calculation resource state of the intelligent terminal 21, planning a local calculation task and unloading calculation tasks according to the wireless bandwidth resource state of the intelligent terminal 21, and designating a corresponding access base station for each unloading calculation subtask when the terminal is accessed to a plurality of base stations, wherein the mobile terminal part decision information unit correctly reorganizes the calculation subtask results returned in disorder;
the mobile communication network part decision information unit is configured to allocate corresponding access points for unloading calculation task data and returning result data to complete receiving and sending according to a wireless bandwidth state between the intelligent terminal 21 and each access base station and a backhaul network congestion state of each base station, and specifically includes:
when the terminal application wants to initiate a calculation unloading request remotely, informing a corresponding base station to be responsible for receiving air interface data containing a corresponding remote calculation subtask from the terminal;
when the distributed cloud computing node returns a sub-computing task result, a corresponding base station is assigned to send air interface data containing the returned sub-computing task result to a designated terminal;
considering the mobility of users, the request of the same sub-calculation task and the sending and receiving base stations of the result need to be respectively assigned dynamically; when the unloading calculation task data arrives at the core network of the mobile network 22, a route is selected for each unloading calculation subtask data according to the state (topology, link flow and node load) of the core network, so that the efficient forwarding of each unloading calculation subtask data to the designated distributed cloud computing node is realized, and the network load and time delay of the core network are reduced;
conversely, when the distributed cloud computing node returns the task results of each unloading computing sub-task, route forwarding to the assigned access network node is provided; the process can realize the dynamic adaptation of the route forwarding plane by means of a software defined network technology;
the distributed cloud computing node decision information unit is used for assigning corresponding distributed cloud computing nodes for each unloading computing subtask to complete computing bearing tasks according to the state information of each distributed cloud computing node; the method comprises the steps that each distributed cloud computing node is used for deciding that each distributed cloud computing node bears virtualized resources required by corresponding unloading computing sub-tasks, and a virtual machine is generated to complete the computing tasks; the parallel collaboration is realized for the computing processes among a plurality of distributed cloud computing nodes.
As shown in fig. 5-8, the heterogeneous distributed cloud platform 1 comprises a heterogeneous distributed artificial intelligent cloud computing center 11, a distributed deep learning platform 12, a deep learning large-scale training system 13, a heterogeneous super computing platform 14 and a heterogeneous basic algorithm library 15;
the heterogeneous distributed artificial intelligence cloud computing center 11 is used for realizing artificial intelligence real-time interaction between a 5G network and a cloud; the distributed deep learning platform 12 is used for supporting the custom extension of a deep learning model and an algorithm and supporting the CPU-GPU or GPU-GPU or CPU-FPGA-GPU hybrid distributed operation.
The deep learning large-scale training system 13 is used for multi-machine multi-CPU-FPGA-GPU hybrid distributed deep learning model training, supports a model with trillion-level parameters, supports the large-scale classification of billions of categories, adopts advanced memory optimization and communication optimization technology, and greatly improves the speed of company training and iterative model by hundreds of CPU-FPGA-GPU hybrid distributed joint training, as shown in FIG. 6;
the heterogeneous supercomputer platform 14 provides multiple computing clusters, central unified storage, lightweight virtualization and continuous computing power support for researchers, as shown in fig. 7;
the heterogeneous basic algorithm library 15 stores various machine learning algorithms including deep neural networks and mathematical and image processing algorithms; compared with an industry open source platform library, the performance of the method is improved by 2-5 times. The system supports mainstream cloud, personal computers, mobile terminals and embedded-end hardware platforms, and supports various system platforms such as Linux, android, iOS and Windows, etc., as shown in FIG. 8.
The implementation mode specifically comprises the following steps: the heterogeneous distributed cloud platform 1 supports the distributed deep learning platform 12 with the customized extension of the deep learning model and algorithm, so that huge load and energy consumption brought by the traditional mobile cloud computing to a Wide Area Network (WAN) can be avoided;
the 5G technology is integrated through the heterogeneous distributed artificial intelligent cloud platform computing center, intelligent collaborative innovation integration of a remote AI and a terminal side AI is enabled to be possible, real-time interaction is carried out through a 5G network and a cloud, data processing capacity is improved, time delay is reduced, the bottleneck of an AI transmission technology is broken through by the fused 5G technology, and intelligent energization is achieved;
in the 5G era, the invention adopts the high-performance heterogeneous distributed cloud platform 1, combines the deep learning large-scale training system 13, and provides a powerful enterprise artificial intelligence innovation service solution in the 5G intelligence era for users by utilizing a heterogeneous high-performance computing center and a high-performance heterogeneous basic algorithm library 15.
The last points to be described are: first, in the description of the present application, it should be noted that, unless otherwise specified and defined, the terms "mounted," "connected," and "connected" are to be construed broadly, and may be mechanical or electrical, or may be a direct connection between two elements, and "upper," "lower," "left," "right," etc. are merely used to indicate relative positional relationships, which may be changed when the absolute position of the object being described is changed;
secondly: in the drawings of the disclosed embodiments, only the structures related to the embodiments of the present disclosure are referred to, and other structures can refer to the common design, so that the same embodiment and different embodiments of the present disclosure can be combined with each other under the condition of no conflict;
finally: the foregoing description of the preferred embodiments of the invention is not intended to limit the invention to the precise form disclosed, and any such modifications, equivalents, and alternatives falling within the spirit and principles of the invention are intended to be included within the scope of the invention.
Claims (9)
1. An artificial intelligence cloud platform towards 5G age, its characterized in that: the heterogeneous distributed cloud platform (1) takes a distributed mobile cloud computing cooperative framework (2) as a basic framework; the distributed mobile cloud computing collaborative architecture (2) comprises an intelligent terminal (21), a mobile network (22), a distributed mobile cloud computing server (23), a collaborative controller (24) and an end-to-end computing offloading service quality assurance mechanism system (25), wherein the distributed mobile cloud computing server (23) is connected with a service base station (3), and the collaborative controller (24) is connected with the collaborative server; the intelligent terminal (21) is used as an initiating terminal of mobile cloud computing, periodically uploads state sensing information of the intelligent terminal through a mobile communication network and receives relevant unloading segmentation decisions calculated by the cooperative controller (24); the mobile network (22) provides wireless access and transmission for an intelligent terminal (21) initiating a computational offload request; the distributed mobile cloud computing server (23) is deployed on a small server or a multi-small server cluster at a mobile access network side, the load state and the virtual machine computing capacity perception information are periodically uploaded to a cooperative server, and a cooperative controller (24) is received to make decision information to reserve virtual machine resources for computing and unloading tasks; the cooperative controller (24) is used for collecting perception information of the intelligent terminal (21), the mobile network (22) and the distributed mobile cloud computing server (23), generating a calculation unloading segmentation decision and transmitting the calculation unloading segmentation decision to the intelligent terminal (21), and transmitting a resource reservation decision to the service base station (3) and the distributed mobile cloud computing server (23); the end-to-end-terminal computing offloading service quality assurance mechanism system (25) comprises a distributed cloud computing perception module and a collaborative decision-making module, wherein the distributed cloud computing perception module respectively works at a server level and a virtual machine level, and the collaborative decision-making module comprises a mobile terminal part decision-making information unit, a mobile communication network part decision-making information unit and a distributed cloud computing node decision-making information unit; the heterogeneous distributed cloud platform (1) comprises a heterogeneous distributed artificial intelligent cloud computing center (11), a distributed deep learning platform (12), a deep learning large-scale training system (13), a heterogeneous super computing platform (14) and a heterogeneous basic algorithm library (15); the deep learning large-scale training system (13) is used for multi-machine multi-GPU distributed deep learning model training, supports a model with trillion-level parameters and carries out large-scale classification of hundreds of billions; the heterogeneous supercomputing platform (14) provides a plurality of computing clusters, central unified storage, lightweight virtualization and continuous computing capability support for researchers; the heterogeneous basic algorithm library (15) stores various machine learning algorithms and mathematical and image processing algorithms including a deep neural network; the heterogeneous distributed artificial intelligence cloud computing center (11) is used for realizing artificial intelligence real-time interaction between a 5G network and a cloud; the distributed deep learning platform (12) is used for supporting the custom extension of a deep learning model and an algorithm and supporting a large number of general CPUs, GPUs or CPU and GPU mixed distributed operations.
2. The 5G age oriented artificial intelligence cloud platform of claim 1, wherein: the co-controllers (24) are in the form of instances or virtual machines running in distributed cloud computing servers or other network elements coexisting with operators, in particular service gateways, packet data gateways and policy and resource management modules.
3. The 5G age oriented artificial intelligence cloud platform of claim 1, wherein: the intelligent terminal (21) is connected with a mobility management entity through a local gateway, the mobility management entity is connected with a packet data gateway and a strategy and resource management equipment through a service gateway, the packet data gateway and the strategy are connected with the resource management equipment, the packet data gateway and the strategy are respectively connected with an operation business terminal and the Internet, and the intelligent terminal (21) is connected with a distributed mobile cloud computing server (23) through the local gateway.
4. The 5G age oriented artificial intelligence cloud platform of claim 1, wherein: the distributed cloud computing perception module is used for collecting the load condition of the whole computing node server or the server cluster at a server level, and specifically comprises server throughput, server concurrency communication state, server computing resource use condition, server storage resource occupancy rate and when the computing node is the server cluster and a virtualization technology is adopted in the cluster to realize a virtualization resource pool of the whole cluster, the server level perception information fully considers the whole condition of the virtual resource pool.
5. The 5G age oriented artificial intelligence cloud platform of claim 1, wherein: the distributed cloud computing perception module is used for collecting virtual machine state information in the whole node at the virtual machine level, and the information comprises: the number of virtual machines, the amount of computing and storage resources occupied by each virtual machine, throughput produced by each virtual machine, and status information related to the bearer resources.
6. The 5G age oriented artificial intelligence cloud platform of claim 1, wherein: the collaborative controller (24) fully grasps state information of the intelligent terminal (21), the mobile network (22) and each distributed cloud computing node by collecting perception information of the intelligent terminal (21), the mobile communication network and the distributed mobile cloud computing server (23), and the collaborative decision module generates collaborative decisions and respectively issues the collaborative decisions to the intelligent terminal (21), the mobile network (22) network element and the distributed cloud computing node by comprehensively analyzing the grasped information, and each part executes respective corresponding actions according to the decisions.
7. The 5G age oriented artificial intelligence cloud platform of claim 1, wherein: the mobile terminal part decision information unit is used for determining the calculation subtask division of corresponding mobile application according to the current battery, energy consumption and calculation resource state of the intelligent terminal (21), planning a local calculation task and unloading calculation tasks according to the wireless bandwidth resource state of the intelligent terminal (21), and when the terminal is accessed to a plurality of base stations, designating a corresponding access base station for each unloading calculation subtask, wherein the mobile terminal part decision information unit correctly reorganizes the calculation subtask results returned in disorder.
8. The 5G age oriented artificial intelligence cloud platform of claim 1, wherein: the mobile communication network part decision information unit is used for distributing corresponding access points for unloading calculation task data and returning result data to finish receiving and sending according to the wireless bandwidth state between the intelligent terminal (21) and each access base station and the feedback network congestion state of each base station.
9. The 5G age oriented artificial intelligence cloud platform of claim 1, wherein: the distributed cloud computing node decision information unit is used for assigning corresponding distributed cloud computing nodes for each unloading computing subtask to complete computing bearing tasks according to the state information of each distributed cloud computing node; the method comprises the steps that each distributed cloud computing node is used for deciding that each distributed cloud computing node bears virtualized resources required by corresponding unloading computing sub-tasks, and a virtual machine is generated to complete the computing tasks; the parallel collaboration is realized for the computing processes among a plurality of distributed cloud computing nodes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910426197.8A CN111984364B (en) | 2019-05-21 | 2019-05-21 | Artificial intelligence cloud platform towards 5G age |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910426197.8A CN111984364B (en) | 2019-05-21 | 2019-05-21 | Artificial intelligence cloud platform towards 5G age |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111984364A CN111984364A (en) | 2020-11-24 |
| CN111984364B true CN111984364B (en) | 2023-05-26 |
Family
ID=73436212
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910426197.8A Active CN111984364B (en) | 2019-05-21 | 2019-05-21 | Artificial intelligence cloud platform towards 5G age |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111984364B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112637290B (en) * | 2020-12-14 | 2024-03-19 | 厦门宏泰科技研究院有限公司 | Global communication network system based on micro base station and edge calculation |
| CN112804446B (en) * | 2020-12-30 | 2022-08-12 | 江苏达科信息科技有限公司 | Big data processing method and device based on cloud platform big data |
| CN112512069B (en) * | 2021-02-02 | 2021-05-28 | 网络通信与安全紫金山实验室 | Network intelligent optimization method and device based on channel beam pattern |
| CN113067873B (en) * | 2021-03-19 | 2022-08-30 | 北京邮电大学 | Edge cloud collaborative optimization method based on deep reinforcement learning |
| CN113301141B (en) * | 2021-05-20 | 2022-06-17 | 北京邮电大学 | Deployment method and system of artificial intelligence support framework |
| CN113316159B (en) * | 2021-05-27 | 2022-09-02 | 北京邮电大学 | 5G network system based on heterogeneous physical layer |
| CN113473449A (en) * | 2021-07-02 | 2021-10-01 | 深圳市威宇智通科技有限公司 | Intelligent connection system based on Internet of things terminal |
| CN116208975A (en) * | 2021-11-30 | 2023-06-02 | 华为技术有限公司 | Task processing method and device |
| CN115002038B (en) * | 2022-08-04 | 2022-10-25 | 中国电子科技集团公司第十五研究所 | Intelligent peak regulation method and system based on cloud distributed coordination service |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102681899A (en) * | 2011-03-14 | 2012-09-19 | 金剑 | Virtual computing resource dynamic management system of cloud computing service platform |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9626227B2 (en) * | 2015-03-27 | 2017-04-18 | Intel Corporation | Technologies for offloading and on-loading data for processor/coprocessor arrangements |
| US10044798B2 (en) * | 2016-02-05 | 2018-08-07 | International Business Machines Corporation | Context-aware task offloading among multiple devices |
| US10747568B2 (en) * | 2017-05-30 | 2020-08-18 | Magalix Corporation | Systems and methods for managing a cloud computing environment |
| US20190095245A1 (en) * | 2017-09-22 | 2019-03-28 | CloudHealth Technologies, Inc. | System and Method for Apportioning Shared Computer Resources |
| US10671435B1 (en) * | 2017-10-19 | 2020-06-02 | Pure Storage, Inc. | Data transformation caching in an artificial intelligence infrastructure |
-
2019
- 2019-05-21 CN CN201910426197.8A patent/CN111984364B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102681899A (en) * | 2011-03-14 | 2012-09-19 | 金剑 | Virtual computing resource dynamic management system of cloud computing service platform |
Non-Patent Citations (4)
| Title |
|---|
| Cloud computing for deep learning analytics: A survey of current trends and challenges;Saiyeda, Anam et al.;International Journal of Advanced Research in Computer Science;第08卷(第02期);第68-72页 * |
| X. Wei et al..MVR: An architecture for computation offloading in mobile edge computing.2017 IEEE International Conference on Edge Computing (EDGE).2017,第232-235页. * |
| 基于安全保障的边缘计算卸载方案;廉晓飞等;中兴通讯技术;第25卷(第02期);第41-46+56页 * |
| 面向55G的分布式移动云计算协同架构与管理机制;李宏佳;陈鑫;周旭;;中兴通讯技术(02);第14-19页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111984364A (en) | 2020-11-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111984364B (en) | Artificial intelligence cloud platform towards 5G age | |
| Qiao et al. | Collaborative task offloading in vehicular edge multi-access networks | |
| Islam et al. | A survey on task offloading in multi-access edge computing | |
| Tzanakaki et al. | Wireless-optical network convergence: Enabling the 5G architecture to support operational and end-user services | |
| Guo et al. | Computation offloading for multi-access mobile edge computing in ultra-dense networks | |
| Baccarelli et al. | Fog of everything: Energy-efficient networked computing architectures, research challenges, and a case study | |
| CN109951821B (en) | Task unloading scheme for minimizing vehicle energy consumption based on mobile edge calculation | |
| Zhou et al. | SAGECELL: Software-defined space-air-ground integrated moving cells | |
| Xiao et al. | Dynamic network slicing for scalable fog computing systems with energy harvesting | |
| Jararweh et al. | The future of mobile cloud computing: integrating cloudlets and mobile edge computing | |
| Stojmenovic | Fog computing: A cloud to the ground support for smart things and machine-to-machine networks | |
| Wang et al. | AI-based cloud-edge-device collaboration in 6G space-air-ground integrated power IoT | |
| CN112333717A (en) | 5G access network slice resource allocation method and device considering power multi-service requirements | |
| Alnoman et al. | Energy efficiency on fully cloudified mobile networks: Survey, challenges, and open issues | |
| Maiti et al. | An effective approach of latency-aware fog smart gateways deployment for IoT services | |
| Park et al. | Network intelligence based on network state information for connected vehicles utilizing fog computing | |
| Ali et al. | Mobile edge computing: A promising paradigm for future communication systems | |
| Meng et al. | Integration application of 5g and smart grid | |
| Haitao et al. | Multipath transmission workload balancing optimization scheme based on mobile edge computing in vehicular heterogeneous network | |
| Gadasin et al. | Organization of Interaction between the Concept of Fog Computing and Segment Routing for the Provision of IoT Services in Smart Grid Networks | |
| Sun et al. | Harmonizing artificial intelligence with radio access networks: Advances, case study, and open issues | |
| Song et al. | Multi-task and multi-objective joint resource optimization for UAV-assisted air-ground integrated networks under emergency scenarios | |
| Venkateswararao et al. | Binary-PSO-based energy-efficient small cell deployment in 5G ultra-dense network | |
| Yu et al. | UAV-assisted cooperative offloading energy efficiency system for mobile edge computing | |
| Ibrahim et al. | Radio access network techniques beyond 5G network: A brief overview |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |