CN111432005A - Service migration method under narrow-band weak networking condition - Google Patents
Service migration method under narrow-band weak networking condition Download PDFInfo
- Publication number
- CN111432005A CN111432005A CN202010234294.XA CN202010234294A CN111432005A CN 111432005 A CN111432005 A CN 111432005A CN 202010234294 A CN202010234294 A CN 202010234294A CN 111432005 A CN111432005 A CN 111432005A
- Authority
- CN
- China
- Prior art keywords
- service
- node
- edge
- migration
- access
- 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.)
- Granted
Links
Images
Classifications
-
- 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
- H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
- H04L67/63—Routing a service request depending on the request content or context
-
- 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
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mobile Radio Communication Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to a service migration method under the condition of a narrow-band weak networking, which comprises the following steps: (a) enabling each edge service node and the end user served by the edge service node to monitor the state of the logical connection with each other through periodic beacon information; (b) when the logic connection fails, the edge service node reports migration requirements to a virtual main node in a cloud center or a mobile edge service node set, and the end user actively searches other possible access objects and delegates the access objects to report to the cloud center or the virtual main node after successful access; (c) and the cloud center or the virtual main node determines a migration path and migrates the migration path according to the maintained and updated cloud-edge or edge-edge fusion network routing graph after acquiring the addresses of the initiating node of the service to be migrated and the receiving node expected to receive the migration service. Therefore, the migration efficiency of the service and the data is improved, and the application service is ensured to be uninterrupted.
Description
Technical Field
The invention belongs to the technical field of network services, relates to a method for service migration under the condition of a narrow-band weak networking network, and particularly relates to a method for improving the service migration and data efficiency under the conditions of intermittent disconnection and narrow-band weak networking.
Background
Zhang Chunming proposes a duplicate consistency model RCMTS facing mass data and based on a time stamp, the RCMTS model manages duplicates through a time stamp technology, and divides duplicate update into two strategies of update outside a domain and update inside the domain through the characteristic of high autonomy of grid regionality, thereby improving the update speed, and adopts an access strategy based on user view, thereby ensuring the correctness of a user accessing a logic file, and simultaneously provides a dynamic extensible duplicate positioning method DSR L, uses an index information node to support the simultaneous efficient positioning of a plurality of duplicates of the same data, and uses a local index node to support the query of the local duplicates, thereby further providing a dynamic mapping technology which can distribute the global duplicate positioning information according to the machine performance of the index node, and support the dynamic adding and quitting of the index information node.
However, when the user equipment performs cross-node access, there are two uninterrupted service provisioning modes for ensuring uninterrupted service. One is the idea of container-based lightweight service and data unitization, which performs service migration with users as the center, and aims to migrate services and data to an edge service node close to user equipment to obtain nearby services; and the other is to establish a relay channel between the new access point and the original access point, bridge the new access point and the original access point, and use the new access point as a relay between the user and the original access point. The former is advantageous to avoid traffic response timeouts, but service and associated data migration times may be longer. The latter, although not requiring migration services, has a service response time that is limited by the relay forwarding capability. Both of these two security modes based on migration face the problem of how to improve the migration service and data efficiency under the conditions of intermittent disconnection and narrow-band weak networking.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a method for service migration under the condition of a narrow-band weak networking network.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows: a method for service migration under a narrowband weak networking condition comprises the following steps:
(a) enabling each edge service node and the end user served by the edge service node to monitor the state of the logical connection with each other through periodic beacon information;
(b) when the logic connection fails, the edge service node reports migration requirements to a virtual main node in a cloud center or a mobile edge service node set, and the end user actively searches other possible access objects and delegates the access objects to report to the cloud center or the virtual main node after successful access;
(c) and the cloud center or the virtual main node determines a migration path and migrates the migration path according to the maintained and updated cloud-edge or edge-edge fusion network routing graph after acquiring the addresses of the initiating node of the service to be migrated and the receiving node expected to receive the migration service.
Optimally, in the step (a), the access state information of the end user is periodically reported to the virtual master node in the cloud center or the mobile edge service node set through the edge service node accessed by the end user.
Further, in step (a), the access status information includes an end user identification number and an end user current location, an edge service node identification number and an edge service node current location of access, and a number of successful accesses, and the information content of the number of successful accesses includes a logical connection maintaining time of each access and is listed in the order of the access request time.
Further, in step (a), the cloud center or the virtual master node calculates the access request density and the measure of the average access maintaining time for each end user, and represents the access request density and the measure with a historical access time sequence diagram; the request density is the frequency of finding access objects and sending requests in a given time length to measure the frequency of one end user to encounter a certain edge service node, and the average access maintenance time is the average value of all successfully accessed connection maintenance time lengths in the given time length to measure the stability of connection.
Further, in the step (a), the cloud center or the virtual master node performs clustering based on a k-means clustering algorithm by using the number of edge service nodes as a parameter k, edge service node coordinates as an initial clustering center, and a distance between an end user and the clustering center as a metric value, so as to generate k clusters.
Optimally, in the step (c), the migration adopts a service lightweight containerization package migration technology.
Due to the application of the technical scheme, compared with the prior art, the invention has the following advantages: according to the method for service migration under the condition of the narrow-band weak networking, the edge service node is connected with the end user through monitoring logic, when the edge service node has a migration requirement, the cloud center or the virtual main node determines the addresses of an initiating node of a service to be migrated and a receiving node which is expected to receive the migration service, and migration is carried out according to the routing graph of the converged network, so that migration efficiency of the service and data is improved, and uninterrupted application service is ensured.
Drawings
FIG. 1 is a schematic structural view of a container mirror layering framework of the present invention;
FIG. 2 is a diagram of the user, service and data interdependencies of the present invention;
FIG. 3 is a diagram of data rights management based on user, service, and data interdependencies according to the present invention;
FIG. 4 is a diagram of data unit and lightweight service migration in accordance with the present invention;
fig. 5 is a frame diagram of an intelligent decision making process for cross-node access and migration in a wireless communication environment.
Detailed Description
Preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings.
The invention relates to a method for service migration under the condition of a narrow-band weak networking, which comprises the following steps:
(a) each edge service node and the end users served by the edge service node are enabled to monitor the state of the logical connection with each other through periodic beacon information.
The access state information of each end user is periodically reported to a virtual main node in a cloud center (cloud for short) or a mobile edge service node (edge for short) set through an edge service node accessed by the end user. The type state information comprises an end user identification number (or address) and a current position, an edge service node identification number (or address) and a current position of access, and a successful access number (comprising a logical connection maintenance time of each access, listed according to a time sequence of request access). Based on such state information, the cloud center or virtual master node calculates the following metric values for each end user: access request density (the number of times of finding an access object and sending a request in a given time length to measure the frequency of an end user accidentally encountering a certain edge service node), average access maintenance time (the average value of all successfully accessed connection maintenance time lengths in the given time length to measure the stability of connection), and a historical access time sequence diagram representation, namely, an abscissa value is a state information reporting time, and an ordinate value is two state values.
The method comprises the steps that virtual main nodes in a cloud center or a mobile edge service node set are clustered on the basis of a k-means clustering algorithm by taking the number of edge service nodes as a parameter k, edge service node coordinates as an initial clustering center and the distance between an end user and the clustering center as a measurement value, and k clusters are generated. The distance value between the edge service node position in the cluster and the central point is divided into three levels of near, middle and far, the number of devices in the inner end of the cluster is divided into three levels of small, middle and large, and the potential access load of the edge service node is comprehensively measured by the two dimensions. The 9 potential access load levels can be formed by combination, for example, if the number of end users of a cluster in which an edge service node which is close to the center of the cluster is located is "large", the potential access load of the edge service node may be the heaviest. The potential access load of each edge service node is also represented as a potential load timing graph.
Based on the two types of timing diagrams, a decision can be made for service migration. For example, after determining the source edge service node and the destination edge service node to be migrated, whether to actually implement the migration needs to be determined as follows: firstly, a potential load sequence diagram of a destination edge service node is consulted, and if the load is judged to be large, the migration service is not suitable to be received; if the judgment can be received, looking up a historical access sequence diagram of a service object (namely a certain end user) of the migration service about a target edge service node, and if the judgment shows that the request access density is extremely low or the average access maintenance time is extremely short, the migration is not suitable; and if the migration is judged to be appropriate, looking up a historical access sequence diagram of the end user about the source edge service node, and if the request access density is judged to be extremely low or the average access maintenance time is extremely short, not considering the retained service copy. In order to improve the accuracy and timeliness of the time series graph analysis, a principal component analysis method, an independent component analysis method, a principal feature analysis method, and a logistic regression method may be used for combined analysis.
(b) When the logic connection fails, the edge service node reports migration requirements to a virtual main node in a cloud center or a mobile edge service node set, and a terminal user actively searches other possible access objects and delegates the access objects to report to the cloud center or the virtual main node after successful access;
(c) and the cloud center or the virtual main node determines a migration path and migrates the migration path according to the maintained and updated cloud-edge or edge-edge fusion network routing graph after acquiring the addresses of the initiating node of the service to be migrated and the receiving node expected to receive the migration service.
If the cloud center or the virtual master node taking over the cloud center can provide insight into the access behavior trajectory of the end device (i.e., the end user), it is beneficial to make a reasonable retention strategy of the migrated service copy and a general basic service cache deployment strategy. And determining a differentiated part of the service to be migrated according to the running environment condition of the service at the migration destination, reasonably dividing the differentiated part into basic units suitable for transmission, and piggybacking the basic units in the IP packet flow of the equidirectional transmission path to reduce the network bandwidth occupation. If the service copy is determined to be reserved, slimming processing is carried out to reduce resource occupation, and the slimming processing comprises splitting processing of the service, recording a service component list and an assembly flow, reserving a differentiated component and deleting a common component which can be locally copied.
The aim of the light service weight is to reduce the transmission amount during the migration so as to adapt to the intermittent disconnection and narrow-band weak networking environment. Meanwhile, the method should have a fast start feature so as to quickly recover the application service suspended due to migration. By using a container technology, researching a mirror image layering mechanism and a basic environment mirror image preloading strategy, and adopting a service lightweight containerization package migration scheme (namely, a service lightweight containerization package migration technology, as shown in fig. 4) under the conditions of high dynamic and intermittent disconnection of an edge service node set and a narrow-band weak networking, the network bandwidth consumption of the service mirror image is reduced, the migration speed is increased, and the service is quickly recovered.
The software product developed based on the container technology and capable of being used in a deliverable way can be regarded as a container application service mirror image, namely an application program constructed based on a hierarchical relationship, and sequentially comprises a boot file system (such as bootfs), a root file system (such as rootfs), development environment tools (such as emacs and apache) and the application program from bottom to top. Each image layer has a corresponding json file whose purpose is to provide information on what processes should be run on top of the image, what environment variables should be configured for the processes, etc. If the container application service image is to be deployed and provided with a service, the container daemon is required to create a corresponding container application service container according to the image json file, that is, a dynamic container is generated according to a static image. When the container is started, a container layer is newly added on the top of the container application service mirror image, one or more processes are operated, and each process occupies corresponding memory, CPU, virtual network equipment resources and file system resources provided by the mirror image layer file of the container mirror image. The process in the container writes a task file according to a Copy-on-Write mechanism, namely, the file is firstly copied from the container mirror image layer to the uppermost container layer, and then the related process writes the Copy in the container layer. It can be seen that the new data or modified data is stored in the uppermost container layer, while the data of the mirror layer remains unchanged. Based on reasonable layering of images, constructing a new application image is simplified into a multiplexing common image, and business logic programs related to the application are added. In combination with the incremental upload and download mechanism of the mirror repository, new applications can be quickly uploaded to the mirror repository and then quickly distributed to other nodes.
The application provides a container layering mechanism, and a common mirror image of a container is divided into a system kernel layer, an operating system layer, a common component layer, a development language layer, a development framework layer and the like from bottom to top. Developers can select any number of mirror image layers from bottom to top on the left side in the image 1 as basic mirror images according to requirements, and develop application service mirror images on the basis of the mirror image layers. This maximizes the utilization of the common image by taking full advantage of the container layering characteristics. For a specific application service container, only the contents of the application logical layer and the service container layer at the uppermost layer of the mirror are specific to the service, and other lower-layer mirrors can be regarded as a common mirror layer supporting the service. Therefore, when an application service needs to be migrated, the adaptation of the common base image of the destination should be known in advance. If the adaptation is complete, only the contents of the application logic layer and the service container layer specific to the service need to be migrated, so as to reduce the transmission amount during migration (the contents of the service container layer may include process recovery information, environment variable configuration requirements, intermediate results of application services, accompanying data, etc., and may be in the form of a json-like file). If the situation of improper configuration exists, incremental synchronization operation can be performed on the public basic mirror image, and pre-migration of the public basic mirror image can be guided based on a big data analysis result of random access behaviors of the end user. While notifying the destination to preload the common base image. If the common base image needs incremental synchronization operation, the migration operation of the contents of the application logic layer and the service container layer should be prioritized to ensure the timely completion of the preloading (namely, the completion before the contents of the application logic layer and the service container layer reach the destination) as much as possible, so as to accelerate the service recovery speed.
In order to meet the requirements of random access of service follow and data carry, the multi-type data to be migrated needs to be effectively managed, necessary data is quickly extracted, and a foundation is laid for quickly migrating and quickly recovering the service. The multi-type data to be migrated includes service container layer process recovery information, environment variable configuration requirements, intermediate results of application services, various types and formats of data to be carried about, and the like, which increases the difficulty of data management. In order to meet the requirements of various service oriented and data migration scenes and information synchronization by taking a data unit as a carrier, the data unit management and on-demand quick extraction technology is researched, and the difficulty of data management is expected to be reduced.
The essence of a data unit is a dynamic association between a user, a service, data and a mount service. The data units can be divided into different types according to the types of the data or the data sets contained in the data units, and the degree of dependence on the data (such as strong dependence, weak dependence and the like) is recorded in the data units. The composition of a data unit may be a single data, a single file, a database table, an object, etc. The data unit also has aggregability and embeddability, for example, a directory composed of a plurality of files, or a plurality of database tables and a data unit set of a plurality of files can form a new data unit. The same file is allowed to be a constituent element of a plurality of data units. The content of a data unit is allowed to change dynamically in the vertical direction, but not in the horizontal direction. If a lateral change is required, only one more data unit based on this data unit can be created. Storage services such as file systems, relational databases, message queues, etc. may also be managed by the data units. The user obtains various application services by accessing the edge information service system. If a random access action occurs, the service needs to be migrated in order for the access point closer to provide service. The migration of multi-type, multi-format and multi-source data related to the method can be contained in one data unit, and the uniform packaging, uniform migration and rapid recovery of services and data are facilitated.
The creation time of the data unit is not limited and can be initiated at any time of the data lifecycle as needed. For example, in the data acquisition period, an acquirer can divide data units according to data sources, data space-time attributes, data acquisition means and other modes; in the data storage period, a storage initiator can divide a data unit according to the destination position of data landing, such as a certain directory, a certain library and a certain table, so that the flexibility of data storage management is increased; in the data processing period, a processor can declare a data set as a data unit in the data processing process, and then apply various operators to the data unit to accelerate the speed of the data processing process; in the data distribution sharing period, the owner performs data distribution, data sharing and data subscription processes in units of data units. Therefore, when the service is required to be migrated, firstly, according to the content characteristics of the application logic layer and the service container layer, the respective appropriate content storage form (for example, the storage form capable of adapting to the quick recovery service) is adopted, and the storage form is determined according to the state mode of the intermediate result of the current service processing; then, necessary accompanying data or data sets are extracted quickly according to application service logic; and finally, creating data units based on the determined various types of data storage forms to bear the various types of migration data.
User, service and data inter-dependencies are shown in FIG. 2. A user with a legitimate identity can gain access to a "non-data access type service" only with an authenticated identity, e.g., send own information, submit perceived environmental information, ask for non-confidential information, etc. However, when it needs to access a "data access type service", it must resort to an authenticated identity and an obtained authorization token. "data access type services" may involve manipulation of data and therefore must be authenticated with identity and authorization to be allowed. First, a certain user identity using a certain "data access type service" is legal and has a right to use the "data access type service", and then the identity of the "data access type service" is also legal and has a right to operate or process a certain type of data (for example, user collected data, service processed data, system stored data, etc.).
As shown in fig. 3, by requesting registration and submitting audit information, after obtaining feedback that the audit is passed, the user has a legal identity in the edge information service system. Based on the legal identity of the user and the application service request submitted by the user, the system grants proper authority for the user after analyzing the application service logic of the user, so that the user can obtain satisfactory service and cannot infringe the authority of other users or the system. As with the user, any application to be deployed into the system is authenticated in advance to ensure traceability of the source. At the same time, it needs to be granted proper data authority to ensure that the data is used safely. The data entering the system is given by the collector or the publisher according to the source of the data, and the data is subjected to authority requirements of being accessed by the application service, and unified authority setting, updating, publishing and other operations by the data authority management service of the system. Aiming at data collected by a user, the user can carry out unitized processing on the data according to the collected time, place, collection means and the like, and correspondingly attaches a data access strategy for reference when the data authority management service carries out processing. In the operation process of the application service, if the generated intermediate result needs to be shared to a related cooperation partner, a data unit can be created based on the shared content, and a data access strategy is correspondingly attached for reference when the data authority management service processes. The data stored in the system is also managed by the data authority management service, data unitization processing is carried out according to the requirements of the application service on data types, data formats, data processing means and the like, reasonable access authority is set, and the safety of the data is guaranteed on the premise that the data access requirements of the application service are met.
Under two network conditions of ensuring network communication and uncertain wireless network conditions, a service and data cross-node access and migration technology is researched, application service characteristics in a maneuvering environment are represented by using characteristics of five dimensions of calculation intensive type, data intensive type, delay sensitive type, fault-tolerant control type, interaction frequent type and the like, and then a technical approach of the service and data cross-node access and migration technology in a typical combination scene is set forth.
The main characteristics of "compute intensive" applications: a large amount of CPU resources need to be consumed; usually, one computing task occupies a large number of computing nodes at the same time; the core number of the CPU is matched with the quantity of the concurrent subtasks; typical applications include high definition video decoding, deep learning training processes, and the like.
The main characteristics of the "data intensive" application: a large number of independent data analysis processing tasks are distributed on different loosely coupled computing nodes for processing; have a highly intensive mass data I/O throughput requirement; typically with a data flow driven flow; typical applications include Web applications, software-as-a-service cloud facilities, information systems with high demands for acquisition, management, analysis, and understanding of data that changes in large quantities and at high speeds, and the like. Such applications need to rely on data intensive computing power, including high performance computing power, data analysis and mining capabilities, and the like.
The main characteristics of the delay sensitive application are as follows: most applications have certain delay requirements and there is usually a tolerable threshold. When this threshold is small, a high priority resource allocation is required to guarantee. Due to the limited resources, the difficulty of guarantee is increased, the probability that effective guarantee cannot be obtained is increased, and the application experience of the user is poor. "delay sensitive" applications refer to applications that require a delay threshold to be set small to meet traffic requirements. Services with the "delay sensitive" feature are also sensitive to temporary interruptions of the logical connection.
The main characteristics of the application of the fault-tolerant control type are as follows: fault tolerance control is defined herein as the ability of an information system to provide service at a desired performance or with an acceptable loss of performance when certain features of the information system fail. Services with the feature of "fault-tolerant control type" will be more tolerant to temporary interruptions of the logical connection.
The main characteristics of the application of the interactive frequent type are as follows: generally, network transmission, disk IO and other IO-intensive tasks are executed, but the consumption of CPU resources is small, most of the tasks wait for IO operations to complete, and typical applications include Web applications and the like.
The application service in the mobile environment may include acquiring comprehensive intelligence data from a cloud center and receiving a task instruction, pushing a complex computing task to the cloud center for execution, deploying a service supported by mass data in the cloud center to provide access nearby on an edge platform, integrating data on a plurality of ends by the edge platform so as to share and comprehensively decide, executing a computation-intensive intelligent training task by the cloud center while making an inference based on a training model by the edge to meet the requirement of a delay-sensitive application, performing coarse-to-fine hierarchical processing on the data to alleviate transmission pressure according to the order of the ends, sides and cloud, and caching the service step by step according to the order of the cloud, sides and ends to reduce access delay.
In view of the characteristics of service application in a mobile environment, the actual service types have rich diversity, and the network conditions have the characteristic of high dynamic change. Therefore, a service and data cross-node access and migration technology adapted to different service characteristics under different network conditions is needed, and an integrated cross-node access and migration mechanism transparent to relevant access requesters is constructed. The method mainly breaks through a transparent cross-node access and migration framework of the self-adaptive service characteristics under the cloud-edge and edge-edge communication conditions and a transparent cross-node access and migration framework of the self-adaptive service characteristics under the uncertain network conditions.
An intelligent decision framework for cross-node access and migration in a wireless communication environment is shown in fig. 5 (an edge node with the highest probability of opportunistic connectivity with other connected subnets is selected to replace a cloud center to perform coordination duty, mainly to facilitate access authentication and meta-information synchronization by using a higher probability of opportunistic connectivity than other edge nodes when an end device spans two unconnected subnets. Under the framework, an edge processing engine arranged on an edge computing platform is responsible for receiving service requests of end equipment. No matter the edge computing platform independently decides to establish service for the end equipment or requests the cloud center to collaborate and establish service, the service request is redirected to a cloud processing engine of the cloud center, and intelligent analysis and decision are carried out by means of mass data support and strong computing capacity of the cloud center so as to obtain a connection maintenance strategy. For example, if the end device leaves the currently accessed edge node and is randomly accessed to other edge nodes, the end device uses service migration or data migration based on edge-edge cooperation, or adopts cloud-edge cooperation service and data migration, and needs to be seized through intelligent decision. The intelligent analysis and decision module works based on a trained machine learning model, which can also be advanced to an edge computing platform to reduce response delay. The end equipment also entrusts the currently accessed edge node to feed back the random access experience to the cloud center. And the cloud center executes a machine learning training decision model according to the received various feedbacks and by combining the service characteristics, the dynamic change of network conditions and the like.
The method comprises a rewarded table and a Q table, wherein the rewarded table and the Q table can be represented as a two-dimensional matrix with states as rows and actions as columns, the value of the rewarded table is the benefit of taking a certain action in a given state, and the value of the Q table records the preference of the action.
The state set in the Q-L earning can be obtained by combining the dimensionalities reflecting the service characteristics, such as a calculation intensive type, a data intensive type, a delay sensitive type, a fault-tolerant control type, an interaction frequent type and the like, and the frequency of cross-node access of end equipment requesting services (such as frequent cross-node access, occasional cross-node access and non-cross-node access).
After the cloud center receives a service request redirected by the edge processing engine, the characteristics of the requested service are obtained through an intelligent analysis and decision module, one of five types of dimensions or a combination of the five types of dimensions is used for representing, mining analysis is carried out according to historical data of access behaviors of the end equipment by using a typical data mining method, the frequency category of cross-node access is obtained, and therefore the current state of the end equipment is obtained. And searching a feasible path based on the trained Q table, converting the current state into a state which can meet the logical connection delay constraint end-to-end transmission application performance goodness of fit, and triggering an action of reaching the satisfied state, namely a connection maintenance strategy expected to be found. The timely feedback of the actual experience of the end equipment and the design of the Q value updating function also relate to the stability of the logical connection delay constraint end-to-end transmission application performance.
In a mobile environment, the probability of the networks being fully interconnected at the same time decreases as the environment deteriorates. The probability of having one or several edge-edge interconnected subnetworks isolated at any one time, or even completely isolated edge nodes, increases. Meanwhile, the isolated edge-edge interconnection sub-network or the completely isolated edge node also shows high dynamic property, and the disconnection is separated when the interconnection is aggregated along with time. On one hand, the network topology dynamic change trend of different mobile models (such as random waypoints, random walk, random direction, random travel, smooth random movement, highway movement and Manhattan grid movement) under different node scale scenes is adopted, the network condition that the cloud edge sub-network and the edge sub-network maintain communication and a modeling opportunity communication probability model are discussed, and the design of an intelligent decision framework for cross-node access and migration under the uncertain wireless environment is guided. On the other hand, a virtual heart jumper mechanism is adopted, and communication information is required to be monitored between each pair of nodes in the cloud-edge fusion network. In order to save network bandwidth and make full use of application service data packets for piggybacking, the aforementioned streaming migration technology based on opportunistic piggybacking and the service-feature-oriented and opportunistic piggybacked meta-information synchronous load minimization technology can also be used to transmit heartbeat information packets. Based on the information of heartbeat information packet interaction, the edge nodes isolated from the cloud can be mutually connected to form an independent connected subnet, and the edge node with the highest probability of opportunistic connection with other connected subnets is selected to replace the role of the cloud center, so that the cross-node random access connection work of the end equipment in the connected subnet is coordinated, namely, in the connected subnet, a route for connecting the current access point and the access point broken due to movement is provided for the end equipment accessed by the cross-node, and the access authentication and the meta-information synchronization are assisted.
The above embodiments are merely illustrative of the technical ideas and features of the present invention, and the purpose thereof is to enable those skilled in the art to understand the contents of the present invention and implement the present invention, and not to limit the protection scope of the present invention. All equivalent changes and modifications made according to the spirit of the present invention should be covered within the protection scope of the present invention.
Claims (6)
1. A method for service migration under narrowband weak networking conditions, comprising the steps of:
(a) enabling each edge service node and the end user served by the edge service node to monitor the state of the logical connection with each other through periodic beacon information;
(b) when the logic connection fails, the edge service node reports migration requirements to a virtual main node in a cloud center or a mobile edge service node set, and the end user actively searches other possible access objects and delegates the access objects to report to the cloud center or the virtual main node after successful access;
(c) and the cloud center or the virtual main node determines a migration path and migrates the migration path according to the maintained and updated cloud-edge or edge-edge fusion network routing graph after acquiring the addresses of the initiating node of the service to be migrated and the receiving node expected to receive the migration service.
2. The method of service migration under narrowband weak networking conditions of claim 1, wherein: in the step (a), the access state information of the end user is periodically reported to a virtual main node in a cloud center or a mobile edge service node set through the edge service node accessed by the end user.
3. The method of service migration under narrowband weak networking conditions of claim 2, wherein: in the step (a), the access state information includes an end user identification number and an end user current position, an accessed edge service node identification number and an edge service node current position, and successful access times, and the information content of the successful access times includes logic connection maintaining time of each access and is listed according to the time sequence of the requested access.
4. A method of service migration in narrowband weak networking conditions according to claim 2 or 3, characterized by: in the step (a), the cloud center or the virtual main node calculates the measurement values of the access request density and the average access maintenance time for each end user, and the measurement values are represented by a historical access time sequence diagram; the request density is the frequency of finding access objects and sending requests in a given time length to measure the frequency of one end user to encounter a certain edge service node, and the average access maintenance time is the average value of all successfully accessed connection maintenance time lengths in the given time length to measure the stability of connection.
5. The method of service migration under narrowband weak networking conditions of claim 2, wherein: in the step (a), the cloud center or the virtual master node performs clustering based on a k-means clustering algorithm by using the number of edge service nodes as a parameter k, edge service node coordinates as an initial clustering center, and a distance between an end user and the clustering center as a metric value, so as to generate k clusters.
6. The method for service migration in narrowband weak networking conditions of claim 1, wherein: in the step (c), the migration adopts a service lightweight containerization package migration technology.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010234294.XA CN111432005B (en) | 2020-03-30 | 2020-03-30 | Service migration method under narrow-band weak networking condition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010234294.XA CN111432005B (en) | 2020-03-30 | 2020-03-30 | Service migration method under narrow-band weak networking condition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111432005A true CN111432005A (en) | 2020-07-17 |
| CN111432005B CN111432005B (en) | 2022-07-29 |
Family
ID=71549890
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010234294.XA Active CN111432005B (en) | 2020-03-30 | 2020-03-30 | Service migration method under narrow-band weak networking condition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111432005B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112751924A (en) * | 2020-12-29 | 2021-05-04 | 北京奇艺世纪科技有限公司 | Data pushing method, system and device |
| CN113489787A (en) * | 2021-07-06 | 2021-10-08 | 北京邮电大学 | Method and device for collaborative migration of mobile edge computing service and data |
| CN113852693A (en) * | 2021-09-26 | 2021-12-28 | 北京邮电大学 | Migration method of edge computing service |
| WO2023016309A1 (en) * | 2021-08-09 | 2023-02-16 | International Business Machines Corporation | Distributed machine learning in edge computing |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015034435A1 (en) * | 2013-09-03 | 2015-03-12 | Nanyang Technological University | A method for managing a data center network |
| CN107864055A (en) * | 2017-10-31 | 2018-03-30 | 云宏信息科技股份有限公司 | The management method and platform of virtualization system |
| CN110275758A (en) * | 2019-05-09 | 2019-09-24 | 重庆邮电大学 | A kind of virtual network function intelligence moving method |
-
2020
- 2020-03-30 CN CN202010234294.XA patent/CN111432005B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015034435A1 (en) * | 2013-09-03 | 2015-03-12 | Nanyang Technological University | A method for managing a data center network |
| CN107864055A (en) * | 2017-10-31 | 2018-03-30 | 云宏信息科技股份有限公司 | The management method and platform of virtualization system |
| CN110275758A (en) * | 2019-05-09 | 2019-09-24 | 重庆邮电大学 | A kind of virtual network function intelligence moving method |
Non-Patent Citations (1)
| Title |
|---|
| 彭军，等: "一种车载服务的快速深度Q学习网络边云迁移策略", 《电子与信息学报》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112751924A (en) * | 2020-12-29 | 2021-05-04 | 北京奇艺世纪科技有限公司 | Data pushing method, system and device |
| CN113489787A (en) * | 2021-07-06 | 2021-10-08 | 北京邮电大学 | Method and device for collaborative migration of mobile edge computing service and data |
| CN113489787B (en) * | 2021-07-06 | 2023-01-17 | 北京邮电大学 | Method and device for collaborative migration of mobile edge computing service and data |
| WO2023016309A1 (en) * | 2021-08-09 | 2023-02-16 | International Business Machines Corporation | Distributed machine learning in edge computing |
| US11770305B2 (en) | 2021-08-09 | 2023-09-26 | International Business Machines Corporation | Distributed machine learning in edge computing |
| CN113852693A (en) * | 2021-09-26 | 2021-12-28 | 北京邮电大学 | Migration method of edge computing service |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111432005B (en) | 2022-07-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111432005B (en) | Service migration method under narrow-band weak networking condition | |
| Liu et al. | A survey on edge computing systems and tools | |
| Torabi et al. | Data replica placement approaches in fog computing: a review | |
| Abdulazeez et al. | Offloading mechanisms based on reinforcement learning and deep learning algorithms in the fog computing environment | |
| CN109743261A (en) | SDN-based container network resource scheduling method | |
| CN107615792B (en) | Management method and system for MTC event | |
| CN113852693B (en) | Migration method of edge computing service | |
| CN112631725A (en) | Cloud-edge-cooperation-based smart city management system and method | |
| Hussain et al. | Fog computing model for evolving smart transportation applications | |
| Núñez-Gómez et al. | HIDRA: A distributed blockchain-based architecture for fog/edge computing environments | |
| da Silva Barbosa et al. | A platform for cloudification of network and applications in the Internet of Vehicles | |
| Taghizadeh et al. | An efficient data replica placement mechanism using biogeography-based optimization technique in the fog computing environment | |
| Dou et al. | Architecture of virtual edge data center with intelligent metadata service of a geo-distributed file system | |
| Wali et al. | Fog Computing: Basics, Key Technologies, Open Issues, And Future Research Directionss | |
| Zhang et al. | An optimal container update method for edge‐cloud collaboration | |
| US20230412471A1 (en) | Edge-based artificial intelligence enablement | |
| Zare et al. | Imperialist competitive based approach for efficient deployment of IoT services in fog computing | |
| Midya et al. | An adaptive resource placement policy by optimizing live VM migration for ITS applications in vehicular cloud network | |
| KR102476271B1 (en) | Method for configuration of semi-managed dht based on ndn and system therefor | |
| US11381468B1 (en) | Identifying correlated resource behaviors for resource allocation | |
| Chang et al. | Adaptive edge process migration for iot in heterogeneous cloud-fog-edge computing environment | |
| Shaik | Resource and Service Management for Fog Infrastructure as a Service | |
| Maheswaran et al. | A language for programming edge clouds for next generation iot applications | |
| Ali et al. | ProKube: Proactive Kubernetes Orchestrator for Inference in Heterogeneous Edge Computing | |
| Tarek et al. | Software-Defined Networks Towards Big Data: A Survey |
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 |