CN115314355B - Deterministic network-based power communication network architecture system and method - Google Patents
Deterministic network-based power communication network architecture system and method Download PDFInfo
- Publication number
- CN115314355B CN115314355B CN202210778564.2A CN202210778564A CN115314355B CN 115314355 B CN115314355 B CN 115314355B CN 202210778564 A CN202210778564 A CN 202210778564A CN 115314355 B CN115314355 B CN 115314355B
- Authority
- CN
- China
- Prior art keywords
- service
- layer
- knowledge
- resource
- scheduling
- 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
- 238000004891 communication Methods 0.000 title claims abstract description 48
- 238000000034 method Methods 0.000 title claims description 28
- 230000006870 function Effects 0.000 claims abstract description 57
- 238000013507 mapping Methods 0.000 claims abstract description 20
- 230000007246 mechanism Effects 0.000 claims abstract description 18
- 238000013468 resource allocation Methods 0.000 claims abstract description 9
- 230000006978 adaptation Effects 0.000 claims abstract description 3
- 230000006399 behavior Effects 0.000 claims abstract description 3
- 238000007726 management method Methods 0.000 claims description 49
- 230000005540 biological transmission Effects 0.000 claims description 30
- 238000004364 calculation method Methods 0.000 claims description 14
- 238000007405 data analysis Methods 0.000 claims description 12
- 230000008901 benefit Effects 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 10
- 230000008713 feedback mechanism Effects 0.000 claims description 7
- 230000001360 synchronised effect Effects 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 4
- 230000003139 buffering effect Effects 0.000 claims description 4
- 238000012423 maintenance Methods 0.000 claims description 4
- 241001522296 Erithacus rubecula Species 0.000 claims description 2
- 230000001419 dependent effect Effects 0.000 claims description 2
- 238000010801 machine learning Methods 0.000 claims description 2
- 239000011159 matrix material Substances 0.000 claims description 2
- 230000008569 process Effects 0.000 claims description 2
- 238000007493 shaping process Methods 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 5
- 230000006855 networking Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 125000004122 cyclic group Chemical group 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000035755 proliferation Effects 0.000 description 1
- 230000010076 replication Effects 0.000 description 1
- 230000000087 stabilizing effect Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/04—Network management architectures or arrangements
- H04L41/044—Network management architectures or arrangements comprising hierarchical management structures
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention provides a deterministic network-based power communication network architecture system, which comprises a control layer, an infrastructure layer and a knowledge layer; the control layer, service and strategy fitting, aiming at the user service requirement, adopts a scheduling mechanism based on knowledge to form a determined route and resource strategy meeting the specific service quality and network behavior characteristic requirement, and realizes decision adaptation of the route and scheduling; an infrastructure layer, wherein the strategy and resource fitting consists of a series of TSN network devices and is used for mapping the routing and scheduling functions into refined network resource combinations so as to realize the forwarding of data; the knowledge layer, the resource and knowledge fit, the communication connection with the infrastructure layer and the control layer, is responsible for collecting and storing the network resource of the infrastructure layer in real time, realizes the dynamic allocation strategy of the network resource by learning the resource allocation strategy of the control layer, and timely feeds back the learned scheduling strategy to the control layer.
Description
Technical Field
The present invention relates to the field of network communication technologies, and in particular, to a deterministic network-based power communication network architecture system and method.
Background
The electric power communication network is used as an important support and guarantee of the power grid, and is an important foundation for realizing the intelligent and interactive control of the power grid and the operation control of the large power grid. With the continuous deep development and promotion of smart power grids and ubiquitous power Internet of things, new technologies such as 5G communication, ICT infrastructure and holographic communication are developed, so that more and more new services are accessed to the network, power communication services are developing to large-bandwidth low-time-delay services such as video, multimedia and accurate load, and new requirements and challenges are provided for the power communication networks. Because the traditional power communication network structure adopts a best effort design mode, the switching mode is single, the transmission delay of a data packet in a network is difficult to determine, and the network controllability is weak. Accordingly, current power systems are urgently required to introduce new technologies and architectures to provide deterministic and low-latency differentiated services.
A deterministic network (Deterministic Network, detNet) is a recently proposed network that guarantees deterministic bandwidth, latency, jitter, and packet loss rate metrics. The method has the advantages that the real-time performance and time certainty of data transmission are guaranteed through the methods of resource reservation, display routing, redundant transmission and the like on the premise of ensuring time accuracy synchronization by the Ethernet TSN technology at the bottom layer. Deterministic network system models have "on-time, reliable, large-scale" properties that are deep-fused with existing power communication networks, and are considered as key technology platforms for grid sensing, computing, and analysis capabilities. The method can match the core requirements of the power grid industry on controllable and differentiated treatment of the service, provide ubiquitous, flexible and efficient brand new technical selection for stabilizing the end-to-end data transmission service for the power access network and the core network, and improve the supporting force of the network architecture.
Therefore, the invention starts from the basic concept of deterministic network and provides a three-layer model of the power communication network integrating deterministic network technology. The role of each layer is elaborated from the aspect of architecture functional design. In order to obtain the real-time network state and autonomously make an optimal decision, the cooperative work of a management control module of the control layer and a management module of the knowledge layer is mainly analyzed, and finally, the real-time information feedback system of the deterministic network is realized.
Disclosure of Invention
The invention aims to: the invention provides an electric power communication network architecture and method based on a deterministic network, which aims to solve the technical problem of effectively reducing the end-to-end time delay of data transmission of the electric power communication network. Due to the specificity of deterministic networks, it is critical to the analysis of their network performance in different application scenarios. The core idea of a Software Defined Network (SDN) is to separate control and forwarding, and to achieve programmability and operational simplicity of management and control. It has been widely used in power communication networks and smart grids. Therefore, based on the idea of SDN control forwarding separation, a three-layer deterministic network architecture similar to SDN is constructed. Wherein the control layer is completely separated from the technical facility layer, and the knowledge layer can communicate with the two layers to timely feed back information.
The technical scheme is as follows: the invention provides an electric power communication network architecture system based on a deterministic network for solving the technical problems, which comprises a control layer, an infrastructure layer and a knowledge layer;
the control layer, service and strategy fitting, aiming at the user service requirement, adopts a scheduling mechanism based on knowledge to form a determined route and resource strategy meeting the specific service quality and network behavior characteristic requirement, and realizes decision adaptation of the route and scheduling;
an infrastructure layer, wherein the strategy and resource fitting consists of a series of TSN network devices and is used for mapping the routing and scheduling functions into refined network resource combinations so as to realize the forwarding of data;
the knowledge layer, the resource and knowledge fit, the communication connection with the infrastructure layer and the control layer, is responsible for collecting and storing the network resource of the infrastructure layer in real time, realizes the dynamic allocation strategy of the network resource by learning the resource allocation strategy of the control layer, and timely feeds back the learned scheduling strategy to the control layer.
Further, the control layer is responsible for realizing the routing and scheduling functions, carrying the service upwards and controlling the data layer downwards, and comprises a data analysis module, a resource management module and a path calculation module, and adopts a scheduling mechanism based on knowledge, and the method comprises the following steps: the data analysis module carries out abstract modeling on the user service request to realize the mapping from the service index to the service model; the path calculation module calculates a deterministic path of the service according to the service model and the routing algorithm; the resource management module adopts different scheduling algorithms according to the service type path and feedback information of the knowledge layer to realize resource allocation among nodes;
wherein the service requirement in the data analysis module comprises a service basic parameter B and a network benefit R, and the basic parameter is defined as five-tuple B= (s, d, l, t) max R), s is a source node, d is a destination node, l is a data stream length, t max R is the distribution rate for the maximum end-to-end delay; the network benefit is defined asWherein v is the service level, c p For operator benefit, c s Is the standard grade gain, delta is the time delay, lambda is jitter, epsilon is the packet loss rateThe method comprises the steps of carrying out a first treatment on the surface of the If the total number of the services requested by the user is m, the basic parameters and the network benefits to be processed by the analysis module are (B) 1 ,B 2 ,...,B m) and (R1 ,R 2 ,...,R m ) In summary, the business requirement model I is expressed as:
the service model S refers to a service performance index and a necessary service function requirement, wherein the performance index is defined as q= (t, δ, λ, epsilon), where t is throughput rate, δ is time delay, λ is jitter rate, epsilon is packet loss rate; the service function requirement F includes: network function F N Functional dependency F Q Network function dependent resource type F R I.e. f= (F N ,F Q ,F R ) The method comprises the steps of carrying out a first treatment on the surface of the When the number of service requests is m, the service model is expressed as:
wherein ,Qi And F is equal to i Respectively representing the performance index of the ith service and the corresponding service function requirement, i epsilon (1, m).
The route addressing function in the path calculation module combines the service requirement matrix S defined above with the routing algorithm to pass through the mapping functionMapping to obtain a service path, wherein the routing algorithm Rt refers to a shortest path algorithm or a K shortest path algorithm, < ->The mapping function is to select all possible transmission paths for the service according to the service performance requirement, and the process of converting the service requirement into the service path by the control layer is expressed as follows:
after the service path is established, carrying out data transmission, and if the application requirement cannot be met or the constraint condition of route adjustment cannot be met, re-executing path calculation;
after the service path meets the conditions, the resource management module adopts a round robin queue forwarding (CQF) mechanism or an asynchronous shaping ATS-based scheduling algorithm to reasonably allocate resources according to the feedback information of the path P and the knowledge layer, so that the service end-to-end delay is minimized and the network resource utilization is maximized.
Further, the infrastructure layer is located at the bottommost layer of the whole architecture and comprises deterministic forwarding equipment and deterministic processing equipment, wherein the deterministic forwarding equipment does not have a routing function and only forwards data; the deterministic processing device has not only data forwarding capability, but also a processing function of realizing data by programming, in particular:
the infrastructure layer functions are implemented by node state information, service capabilities and function instances, where N is defined as all physical nodes N in the network i I is the number of physical nodes; the state information of the node is defined as f= (N l ,N s ,N d ,N im ) The element components in the state information respectively represent node positions, node types, node connection degrees and node importance degrees; service capability is defined as c= (C c ,C h ,C t ) Including computing power C c Caching ability C h Transmission capacity C t Wherein each class of service capability specific division is denoted by a subscript j; function instance e= (E c ,E h ,E t ) Resource service capability of corresponding node, E c To calculate the function instance, E h For caching function instances, E t For a transport function instance, i.e. a service instance of node n on the j-th class of capability of transport resource instance E is defined as E, whereinServices respectively representing nodes on i-th type of capability of computing, buffering and transmitting functionsInstance i e (1, j). Based on the above definition, the resource instantiation result for the physical node on which the kth service node depends on the service path P is:
the mapping from the service path to the data layer service instance resource combination is expressed as:
wherein ,representing a mapping function that converts nodes traversed by path P into refined resource instances in terms of service capability C and service instance F. d, d i And (3) representing the resource instantiation result of the ith service node, i epsilon (1, k).
Further, the knowledge layer is communicated with the infrastructure layer and the control layer through a programmable interface, and comprises an enhanced knowledge management module and a state management information base; wherein, the state management information base stores service resource combination examples, is used for updating the equipment state in real time, enhances the scheduling strategy of the knowledge management module record control layer, combines the machine learning algorithm to learn the scheduling algorithm of different service demands, predicts the optimal scheduling strategy of the data stream according to the learned knowledge, and defines the knowledge as K= (lh, br, D) max ,D T ,C l ) Lh is the data stream length, the data burst rate br, and the maximum end-to-end delay D max Stream deadline D T And link capability C l 。
Further, the control layer cooperates with the knowledge layer to implement a timely and accurate feedback mechanism, including: when the control layer performs data flow scheduling allocation, firstly analyzing the data flow to obtain the performance requirements of the data flow, wherein the performance requirements comprise the tolerable maximum time delay, the length of a data packet and the data burst rate; then, according to the analyzed performance model of the data flow, mapping the route and the service to obtain a deterministic service path; secondly, the control layer requests the existing scheduling knowledge to the knowledge layer, including synchronous scheduling or asynchronous scheduling, and selects a corresponding scheduling strategy according to the bottom network resource state recorded by the information state database; and meanwhile, the knowledge layer carries out intelligent learning decision according to the existing scheduling algorithm and flow characteristics, the learned knowledge is used for predicting the scheduling strategy of the subsequent flow, and finally, the routing and scheduling strategy is issued to the infrastructure layer for corresponding forwarding and processing.
Further, the knowledge management module and the equipment information state database of the knowledge layer collect and update the operation state information of the bottom layer forwarding equipment, and automatically send state update information to the affiliated management module when the operation state of the bottom layer forwarding equipment changes; the enhanced knowledge management module periodically sends state maintenance information to the bottom forwarding equipment so as to prevent the situation that the information cannot be autonomously communicated with the knowledge layer when faults occur, and the control layer directly accesses an information state database when carrying out data flow scheduling distribution to obtain real-time network equipment operation state information, so that a proper strategy is selected for optimal distribution according to service requirements; the enhancement management module is used for excavating and learning the algorithm of the knowledge according to the requirements of the control layer and sending the corresponding knowledge to the control layer, so that intelligent management and decision-making are realized.
The invention also provides a power communication network resource scheduling method realized by the power communication network architecture system based on the deterministic network, which comprises the following steps:
step 1, aiming at a deterministic service of a power grid, a control layer obtains the performance requirement of the service through a data analysis module and maps the performance requirement into a service function model S;
step 2, transmitting the service function model to a route control module, and combining a routing algorithm to realize a deterministic transmission path;
step 3, the resource management module obtains the related knowledge of the resource strategy from the knowledge layer, and invokes a scheduling algorithm to allocate the resources by combining the transmission path P;
step 4, according to the equipment state information recorded by the state information base of the knowledge layer, if the service resource combination meets the routing constraint and the service performance requirement, scheduling is carried out at the moment, the control layer determines that the path allocation resource is allocated to the bottom layer, the execution of a buffering mechanism, a queue scheduling mechanism and a path selection mechanism for the deterministic requirement is completed, and an allocation command is issued to the infrastructure layer; if the scheduling conditions are not met, carrying out scheduling calculation and strategy learning again;
step 5, the infrastructure layer receives the dispatching command of the control layer and transmits the service data packet; after the data stream transmission is completed, the network state is updated, and the state information is stored in a state information base of the knowledge layer.
The beneficial effects are that: compared with the prior art, the invention has the following beneficial technical effects by adopting the technical scheme:
the novel deterministic network-based power communication network architecture and the method adopt a deterministic network architecture based on SDN, are integrated with the existing power network architecture, and realize the requirements of synchronous transmission between substations, low time delay, high reliability and the like of dispatching automation by setting a professional level, applying core technologies such as clock synchronization, display routing and the like.
The novel power communication network architecture and the method based on the deterministic network, which are disclosed by the invention, take deterministic transmission as a core to support reasonable coexistence of multi-deterministic service and common service, provide efficient and convenient management through an efficient queue management mechanism and an information feedback mechanism, realize deterministic service guarantee, and inherit the traditional advantages of optical transmission.
The novel power communication network architecture and method based on the deterministic network have the characteristics of low cost and multiplexing of the Ethernet, and solve the defects that the traditional power architecture is poor in expandability and can not effectively support time-sensitive services and burst services. The deterministic network architecture is used for the comprehensive networking scene of electric power, and provides resources and deterministic service guarantee for business decisions.
Drawings
FIG. 1 is a schematic diagram of a deterministic network-based power communication architecture in accordance with an embodiment of the present invention;
FIG. 2 is a flow chart of a deterministic network-based power communication network communication method in accordance with an embodiment of the present invention;
FIG. 3 is a schematic diagram of a deterministic network based information feedback mechanism according to an embodiment of the present invention;
fig. 4 is a schematic diagram of deterministic network-based management functions according to an embodiment of the present invention.
Detailed Description
The technical scheme of the invention is further described in detail below with reference to the accompanying drawings:
in the current era of data stream proliferation, the existing internet often has the problems of data congestion, high packet loss rate, no guarantee of time delay and the like. The deterministic network provides deterministic service function, and can ensure that data transmission with special requirements can meet extremely low time delay, zero packet loss and high reliable service quality requirements in any network environment. Based on a deterministic network architecture, the method is integrated with the existing communication network architecture, and a solution idea is provided for data transmission of the power communication network.
According to an embodiment of the present invention, a deterministic network-based power communication network architecture, as shown in fig. 1, includes: an infrastructure layer, a control layer, and a knowledge layer.
In particular, the infrastructure layer, which is composed of a series of TSN network devices, is communicatively connected up to the control layer for deterministic forwarding of data. The control layer is composed of a plurality of controllers, logically forms a central controller, is in communication connection with the infrastructure layer and is responsible for controlling and managing the underlying equipment. The knowledge layer is in communication connection with the infrastructure layer and the control layer, is responsible for collecting and storing network resources of the infrastructure layer in real time, realizes the dynamic allocation strategy of the network resources by learning the resource allocation strategy of the control layer, and timely feeds back the learned scheduling strategy to the control layer.
The infrastructure layer and the control layer realize control and forwarding separation, and the knowledge layer, the infrastructure layer and the control layer are mutually communicated to form a dynamic real-time feedback mechanism so as to determine a determined transmission path of the data flow and reasonably allocate resources.
According to some embodiments of the inventionFor example, as shown in fig. 1, the infrastructure layer contains two types of deterministic network nodes, deterministic edge nodes and deterministic forwarding nodes. The deterministic edge node may act as a start point or end point for deterministic flows whose main functions include adding or deleting packet ordering, packet replication and combining. The deterministic forwarding node only realizes a DetNet forwarding sublayer and is responsible for routing the message from a source to a destination and focusing on determining the data forwarding of the network. And when receiving the control information of the control layer, the data forwarding of the service is realized, and meanwhile, the control layer is provided with own state information, so that the control layer is ensured to update the network topology in real time. The control and knowledge plane is also provided with node state and resource use information, and the running state information of the control and knowledge plane is updated in real time so that the control layer can better perform resource allocation and routing. Specifically, the function of the infrastructure layer is to implement a refined resource service combination of routing and resource allocation policies, and three attributes of node state information, service capability and function instance are mainly adopted to implement node instantiation. Node state information is defined as a set of node positions, node types, node connectivity and node importance; the service capability includes computing capability C c Caching ability C h And transmission capability C t The method comprises the steps of carrying out a first treatment on the surface of the The function instance corresponds to the resource service capability of the node; the above attributes pass through mapping functionAnd instantiating the resources of the physical nodes to obtain a service instance resource combination for the knowledge layer to access and store.
According to some embodiments of the invention, as shown in fig. 1, the control layer is composed of a plurality of controllers, actually forming a centralized logic controller. Each controller consists of a plurality of control modules, including a resource management module, a path calculation module, a flow scheduling module and a data analysis module, wherein each module cooperates to control and manage the bottom layer equipment, and the network state can be obtained in time and the corresponding knowledge algorithm can be invoked by interacting with the knowledge plane.
Specifically, the data analysis module carries out abstract modeling on the user service request to realize the mapping from the service index to the service model; the path calculation module calculates a deterministic path of the service according to the service model and a routing algorithm, such as a shortest path algorithm; meanwhile, the resource management module adopts a scheduling algorithm according to the service type path and feedback information of the knowledge layer to realize resource allocation among nodes.
The scheduling algorithm adopted herein can be divided into two modes of synchronous scheduling and asynchronous scheduling. The cyclic forwarding queues based on cyclic forwarding queue or cycle assignment are all synchronous scheduling algorithms, and the network adopts a clock synchronization protocol and a frame preemption mechanism of a TSN time sensitive network to realize quick transmission of delay sensitive flows and ensure extremely low packet loss rate. And an asynchronous traffic shaper ATS adopted by the asynchronous scheduling algorithm is based on an urgent level scheduler (UBS) to realize deterministic transmission per flow. The method is characterized in that strict clock synchronization is not needed, the expandability is strong, and the network bandwidth can be fully utilized.
According to some embodiments of the present invention, as shown in fig. 1, the knowledge layer is composed of an enhanced knowledge management module and a device information status database, and is responsible for collecting and storing network status information, such as the status of flows flowing through each network node. Meanwhile, the existing scheduling algorithm is mined and learned according to the requirements of the control plane, and the network state information and the intelligent learning algorithm can be sent to the control plane according to the requirements of the control plane.
The enhanced knowledge management module is composed of a plurality of devices, each device manages an area and is mainly responsible for communicating with the forwarding nodes of the infrastructure layer, and the operation state information of the forwarding devices of the bottom layer, such as link transmission delay, residual buffer space, available bandwidth, forwarding delay of the end device and the like, is collected and updated. And aggregate the collected information into a device information status database. And each time the running state of the bottom layer forwarding equipment changes, the state updating information is automatically sent to the affiliated management module. In addition, the enhanced knowledge management module periodically sends state maintenance information to the bottom forwarding device so as to prevent the failure from being capable of autonomously communicating with the knowledge layer.
According to the deterministic network-based power communication network networking method provided by the embodiment of the invention, the deterministic network-based power communication network architecture is adopted, a service scheduling method flow chart is shown in fig. 2, and the method comprises the following steps:
(1) The user request arrives, the control layer analyzes the data packet through the data analysis module, so as to obtain the service function requirement of the business, and the service is converted into a service function model;
(2) Transmitting the service function model to a route control module, and realizing a deterministic transmission path P by combining a routing algorithm;
(3) The resource management module obtains the related knowledge of the resource strategy from the knowledge layer, invokes a scheduling algorithm in combination with the transmission path P, and judges the allocation of the resources;
(4) According to the equipment state information recorded by the state information base of the knowledge layer, if the service resource combination meets the routing constraint and the service performance requirement, scheduling is carried out at the moment, the control layer determines that the path allocation resource is allocated to the bottom layer, the execution of a buffering mechanism, a queue scheduling mechanism and a path selection mechanism for the deterministic requirement is completed, and an allocation command is issued to the infrastructure layer; and if the scheduling conditions are not met, carrying out scheduling calculation and strategy learning again.
(5) The infrastructure layer receives a scheduling command of the control layer and transmits the service data packet; after the data stream transmission is completed, the network state is updated, and the state information is stored in a state information base of the knowledge layer.
The hierarchical schematic diagram among the modules of the invention is shown in fig. 1. The system function module comprises: based on the deterministic network architecture of SDN, a deterministic network information feedback mechanism. Deterministic networks require limited latency upper bound and jitter and extremely low packet loss rates, which not only puts higher demands on the forwarding capabilities of the forwarding layer, but also puts more stringent management and control capabilities on the upper layer. In order to realize timely and accurate information feedback, a knowledge management module and a device information state database are enhanced in a knowledge layer, and the knowledge management module and a resource management module of a control layer are cooperatively processed to form a closed-loop information feedback mechanism.
The power communication architecture function of the deterministic network is realized:
the network node adopts deterministic time delay guarantee, the power core networking architecture mainly comprises a knowledge layer, a control layer and an infrastructure layer, a deterministic network architecture and a protocol are deployed in each layer, and the scalability and the collaborative business capability of the deterministic power architecture are completed through collaborative processing of each layer, and a deterministic power communication network architecture schematic diagram is shown in figure 1.
Implementation of the power communication architecture feedback function of deterministic networks, as shown in fig. 3:
the information feedback function is based on the enhancement management module and the deterministic routing protocol, so that information interaction between the knowledge layer and the control layer is realized; based on the information state equipment library, the communication with the infrastructure layer is realized, and the changed network state is timely acquired. And each time the running state of the bottom layer forwarding equipment changes, the state updating information is automatically sent to the affiliated management module. In addition, the enhanced knowledge management module periodically sends state maintenance information to the bottom forwarding device so as to prevent the failure from being capable of autonomously communicating with the knowledge layer. When the control layer performs data flow scheduling allocation, firstly, analyzing the data flow to obtain the performance requirements of the data flow, such as the tolerable maximum time delay, the length of a data packet, the data burst rate and the like; then, according to the analyzed performance model of the data flow, mapping the route and the service to obtain a deterministic service path; secondly, the control layer requests the existing scheduling knowledge from the knowledge layer, such as synchronous scheduling or asynchronous scheduling, and selects a corresponding scheduling strategy according to the bottom network resource state recorded by the information state database; meanwhile, the knowledge layer can carry out intelligent learning decision according to the existing scheduling algorithm and flow characteristics, and the learned knowledge can be used for predicting the scheduling strategy of the subsequent flow. And finally, issuing the routing and scheduling strategy to an infrastructure layer for corresponding forwarding and processing.
Implementation of the power communication architecture management function of deterministic networks, as shown in fig. 4:
the control layer is mainly responsible for the resource and calculation management of the whole framework and mainly comprises a resource management module, a path calculation module, a flow scheduling module and a data analysis module, wherein the modules work cooperatively to control and manage the bottom layer equipment, and the network state can be obtained in time and the corresponding scheduling algorithm can be invoked through interaction with a knowledge plane.
In summary, the invention provides a fusion architecture and method based on a deterministic network for a power communication network system based on an SDN deterministic network architecture, supports large-scale deterministic service business, provides a flexible and extensible architecture, and realizes the functions of low time delay, high-speed movement, high reliability, low packet loss and the like of power communication.
Claims (5)
1. A deterministic network-based power communication network architecture system, characterized in that the system comprises a control layer, an infrastructure layer and a knowledge layer;
the control layer, service and strategy fitting, aiming at the user service requirement, adopts a scheduling mechanism based on knowledge to form a determined route and resource strategy meeting the specific service quality and network behavior characteristic requirement, and realizes decision adaptation of the route and scheduling;
an infrastructure layer, wherein the strategy and resource fitting consists of a series of TSN network devices and is used for mapping the routing and scheduling functions into refined network resource combinations so as to realize the forwarding of data;
the knowledge layer, the resource and knowledge fit, the communication connection with the infrastructure layer and the control layer, the network resource of the infrastructure layer is collected and stored in real time, the dynamic allocation strategy of the network resource is realized by learning the resource allocation strategy of the control layer, and the learned scheduling strategy is fed back to the control layer in time;
the control layer is responsible for realizing the routing and scheduling functions, carrying the service upwards and controlling the data layer downwards, and comprises a data analysis module, a resource management module and a path calculation module, and adopts a scheduling mechanism based on knowledge, and the method comprises the following steps: the data analysis module carries out abstract modeling on the user service request to realize the mapping from the service index to the service model; the path calculation module calculates a deterministic path of the service according to the service model and the routing algorithm; the resource management module adopts different scheduling algorithms according to the service type path and feedback information of the knowledge layer to realize resource allocation among nodes;
wherein the service requirement in the data analysis module comprises a service basic parameter B and a network benefit R, and the basic parameter is defined as five-tuple B= (s, d, l, t) max R), s is a source node, d is a destination node, l is a data stream length, t max R is the distribution rate for the maximum end-to-end delay; the network benefit is defined asWherein v is the service level, c p For operator benefit, c s Is the standard grade gain, delta is the time delay, lambda is the jitter rate, epsilon is the packet loss rate; if the total number of the services requested by the user is m, the basic parameters and the network benefits to be processed by the analysis module are (B) 1 ,B 2 ,...,B m) and (R1 ,R 2 ,...,R m ) In summary, the business requirement model I is expressed as:
the service model S refers to a service performance index and a necessary service function requirement, wherein the performance index is defined as q= (t, δ, λ, epsilon), where t is throughput rate, δ is time delay, λ is jitter rate, epsilon is packet loss rate; the service function requirement F includes: network function F N Functional dependency F Q Network function dependent resource type F R I.e. f= (F N ,F Q ,F R ) The method comprises the steps of carrying out a first treatment on the surface of the When the number of service requests is m, the service model is expressed as:
wherein ,Qi And F is equal to i Respectively representing the performance index of the ith service and the corresponding service function requirement, i epsilon (1, m);
the routing function in the path computation block combines the service requirement matrix S defined aboveRouting algorithm passes mapping functionsMapping to obtain a service path, wherein the routing algorithm Rt refers to a shortest path algorithm or a K shortest path algorithm, < ->The mapping function is to select all possible transmission paths for the service according to the service performance requirement, and the process of converting the service requirement into the service path by the control layer is expressed as follows:
after the service path is established, carrying out data transmission, and if the application requirement cannot be met or the constraint condition of route adjustment cannot be met, re-executing path calculation;
after the service path meets the conditions, the resource management module adopts a round robin queue forwarding (CQF) mechanism or an asynchronous shaping ATS-based scheduling algorithm to reasonably allocate resources according to the feedback information of the path P and the knowledge layer, so that the service end-to-end delay is minimized and the network resource utilization is maximized;
the knowledge layer is communicated with the infrastructure layer and the control layer through a programmable interface and comprises an enhanced knowledge management module and a state management information base; wherein, the state management information base stores service resource combination examples, is used for updating the equipment state in real time, enhances the scheduling strategy of the knowledge management module record control layer, combines the machine learning algorithm to learn the scheduling algorithm of different service demands, predicts the optimal scheduling strategy of the data stream according to the learned knowledge, and defines the knowledge as K= (lh, br, D) max ,D T ,C l ) Lh is the data stream length, the data burst rate br, and the maximum end-to-end delay D max Stream deadline D T And link capability C l 。
2. A deterministic network based power communication network system as defined in claim 1, wherein,
the infrastructure layer is positioned at the bottommost layer of the whole architecture and comprises deterministic forwarding equipment and deterministic processing equipment, wherein the deterministic forwarding equipment has no routing function and only forwards data; the deterministic processing device has not only data forwarding capability, but also a processing function of realizing data by programming, in particular:
the infrastructure layer functions are implemented by node state information, service capabilities and function instances, where N is defined as all physical nodes N in the network i I is the number of physical nodes; the state information of the node is defined as f= (N l ,N s ,N d ,N im ) The element components in the state information respectively represent node positions, node types, node connection degrees and node importance degrees; service capability is defined as c= (C c ,C h ,C t ) Including computing power C c Caching ability C h Transmission capacity C t Wherein each class of service capability specific division is denoted by a subscript j; function instance e= (E c ,E h ,E t ) Resource service capability of corresponding node, wherein E c To calculate the function instance, E h For caching function instances, E t For a transport function instance, i.e. a service instance of node n on the j-th class of capability of transport resource instance E is defined as E, whereinThe service instance i epsilon (1, j) of the node on the i-th class of capability of the computing, caching and transmitting functions is represented respectively, and based on the definition, the resource instantiation result of the physical node on the service path P, on which the kth service node depends, is:
the mapping from the service path to the data layer service instance resource combination is expressed as:
wherein ,representing a mapping function which converts nodes traversed by path P into refined resource instances in terms of service capability C and service instance F, d i And (3) representing the resource instantiation result of the ith service node, i epsilon (1, k).
3. A deterministic network based power communication network system as defined in claim 1, wherein,
the control layer cooperates with the knowledge layer to realize a timely and accurate feedback mechanism, comprising: when the control layer performs data flow scheduling allocation, firstly analyzing the data flow to obtain the performance requirements of the data flow, wherein the performance requirements comprise the tolerable maximum time delay, the length of a data packet and the data burst rate; then, according to the analyzed performance model of the data flow, mapping the route and the service to obtain a deterministic service path; secondly, the control layer requests the existing scheduling knowledge to the knowledge layer, including synchronous scheduling or asynchronous scheduling, and selects a corresponding scheduling strategy according to the bottom network resource state recorded by the information state database; and meanwhile, the knowledge layer carries out intelligent learning decision according to the existing scheduling algorithm and flow characteristics, the learned knowledge is used for predicting the scheduling strategy of the subsequent flow, and finally, the routing and scheduling strategy is issued to the infrastructure layer for corresponding forwarding and processing.
4. A deterministic network based power communication network system as defined in claim 1, wherein,
the knowledge management module and the equipment information state database of the knowledge layer collect and update the operation state information of the bottom layer forwarding equipment, and autonomously send state update information to the affiliated management module when the operation state of the bottom layer forwarding equipment changes; the enhanced knowledge management module periodically sends state maintenance information to the bottom forwarding equipment so as to prevent the situation that the information cannot be autonomously communicated with the knowledge layer when faults occur, and the control layer directly accesses an information state database when carrying out data flow scheduling distribution to obtain real-time network equipment operation state information, so that a proper strategy is selected for optimal distribution according to service requirements; the enhancement management module is used for excavating and learning the algorithm of the knowledge according to the requirements of the control layer and sending the corresponding knowledge to the control layer, so that intelligent management and decision-making are realized.
5. A power communication network resource scheduling method implemented by a power communication network architecture system based on a deterministic network according to any one of claims 1-4, characterized in that the method steps are as follows:
step 1, aiming at a deterministic service of a power grid, a control layer obtains the performance requirement of the service through a data analysis module and maps the performance requirement into a service function model S;
step 2, transmitting the service function model to a route control module, and combining a routing algorithm to realize a deterministic transmission path;
step 3, the resource management module obtains the related knowledge of the resource strategy from the knowledge layer, and invokes a scheduling algorithm to allocate the resources by combining the transmission path P;
step 4, according to the equipment state information recorded by the state information base of the knowledge layer, if the service resource combination meets the routing constraint and the service performance requirement, scheduling is carried out at the moment, the control layer determines that the path allocation resource is allocated to the bottom layer, the execution of a buffering mechanism, a queue scheduling mechanism and a path selection mechanism for the deterministic requirement is completed, and an allocation command is issued to the infrastructure layer; if the scheduling conditions are not met, carrying out scheduling calculation and strategy learning again;
step 5, the infrastructure layer receives the dispatching command of the control layer and transmits the service data packet; after the data stream transmission is completed, the network state is updated, and the state information is stored in a state information base of the knowledge layer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210778564.2A CN115314355B (en) | 2022-06-30 | 2022-06-30 | Deterministic network-based power communication network architecture system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210778564.2A CN115314355B (en) | 2022-06-30 | 2022-06-30 | Deterministic network-based power communication network architecture system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115314355A CN115314355A (en) | 2022-11-08 |
| CN115314355B true CN115314355B (en) | 2023-10-03 |
Family
ID=83855828
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210778564.2A Active CN115314355B (en) | 2022-06-30 | 2022-06-30 | Deterministic network-based power communication network architecture system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115314355B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115632939B (en) * | 2022-12-23 | 2023-03-31 | 浩鲸云计算科技股份有限公司 | Automatic network selection and routing method for achieving multi-target achievement of computational power network |
| CN115632925B (en) * | 2022-12-26 | 2023-03-14 | 北京交通大学 | Time-certainty fusion network architecture and data communication method |
| CN116542000B (en) * | 2023-05-05 | 2024-01-26 | 华能威海发电有限责任公司 | Power grid refinement management system based on source network data analysis |
| CN116819943B (en) * | 2023-08-30 | 2023-11-14 | 浙江大学 | Control system and method capable of realizing task migration flexible function reconstruction |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020181761A1 (en) * | 2019-03-11 | 2020-09-17 | 江苏君英天达人工智能研究院有限公司 | Sdn enhanced path allocation device and method employing bin-packing technique |
| CN112202672A (en) * | 2020-09-17 | 2021-01-08 | 华中科技大学 | Network route forwarding method and system based on service quality requirement |
| CN113079218A (en) * | 2021-04-09 | 2021-07-06 | 网络通信与安全紫金山实验室 | Service-oriented computing power network system, working method and storage medium |
| CN113206796A (en) * | 2021-04-30 | 2021-08-03 | 网络通信与安全紫金山实验室 | Transfer, calculation and storage integrated cooperative system and method |
| CN113435793A (en) * | 2021-08-09 | 2021-09-24 | 贵州大学 | Micro-grid optimization scheduling method based on reinforcement learning |
| US11256238B1 (en) * | 2020-09-10 | 2022-02-22 | Fisher-Rosemount Systems, Inc. | Network resource management in a communication network for control and automation systems |
-
2022
- 2022-06-30 CN CN202210778564.2A patent/CN115314355B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020181761A1 (en) * | 2019-03-11 | 2020-09-17 | 江苏君英天达人工智能研究院有限公司 | Sdn enhanced path allocation device and method employing bin-packing technique |
| US11256238B1 (en) * | 2020-09-10 | 2022-02-22 | Fisher-Rosemount Systems, Inc. | Network resource management in a communication network for control and automation systems |
| CN112202672A (en) * | 2020-09-17 | 2021-01-08 | 华中科技大学 | Network route forwarding method and system based on service quality requirement |
| CN113079218A (en) * | 2021-04-09 | 2021-07-06 | 网络通信与安全紫金山实验室 | Service-oriented computing power network system, working method and storage medium |
| CN113206796A (en) * | 2021-04-30 | 2021-08-03 | 网络通信与安全紫金山实验室 | Transfer, calculation and storage integrated cooperative system and method |
| CN113435793A (en) * | 2021-08-09 | 2021-09-24 | 贵州大学 | Micro-grid optimization scheduling method based on reinforcement learning |
Non-Patent Citations (4)
| Title |
|---|
| Overview and effect of deterministic ethernet on test strategies;Nathan Warden;《2017 IEEE AUTOTESTCON》;全文 * |
| 时间敏感网络中时间触发流路径选择与调度算法研究;陈俊涛;《中国优秀硕士论文电子期刊网》;全文 * |
| 朱瑾瑜等.确定性工业网络架构与关键技术分析.《邮电设计技术》.2022,正文. * |
| 蔡岳平 ; 李天驰 ; .面向算力匹配调度的泛在确定性网络研究.信息通信技术.2020,(第04期),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115314355A (en) | 2022-11-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115314355B (en) | Deterministic network-based power communication network architecture system and method | |
| CN109995583B (en) | Delay-guaranteed NFV cloud platform dynamic capacity expansion and contraction method and system | |
| Garg et al. | MobQoS: Mobility-aware and QoS-driven SDN framework for autonomous vehicles | |
| CN109245916B (en) | Intention-driven cloud access network system and method | |
| CN108322925B (en) | Transmission path calculation method for distinguishing service types in ultra-density heterogeneous convergence network | |
| CN111010295B (en) | SDN-MEC-based power distribution and utilization communication network task migration method | |
| CN105610715B (en) | A kind of cloud data center multi-dummy machine migration scheduling method of planning based on SDN | |
| CN113411223B (en) | Industrial software defined network slicing method based on edge cooperation | |
| CN112995682B (en) | Method and device for deploying and migrating video cloud service | |
| CN113472811B (en) | Heterogeneous service function chain forwarding protocol and method in intelligent fusion identification network | |
| CN114143181B (en) | Intention-driven spatial information network arrangement system and method | |
| US20220150159A1 (en) | Control device, switch device and methods | |
| CN110213175A (en) | A kind of intelligent managing and control system and management-control method towards knowledge definition network | |
| CN113490279B (en) | Network slice configuration method and device | |
| Saravanan et al. | Advance Map Reduce Task Scheduling algorithm using mobile cloud multimedia services architecture | |
| CN112148381A (en) | Software definition-based edge computing priority unloading decision method and system | |
| CN116599904A (en) | Parallel transmission load balancing device and method | |
| CN116755867B (en) | Satellite cloud-oriented computing resource scheduling system, method and storage medium | |
| Saravanan et al. | Improving map reduce task scheduling and micro-partitioning mechanism for mobile cloud multimedia services | |
| Wu | Deep reinforcement learning based multi-layered traffic scheduling scheme in data center networks | |
| CN115604311B (en) | Cloud fusion computing system oriented to service network and self-adaptive routing method | |
| Zhu et al. | Deep reinforcement learning-based edge computing offloading algorithm for software-defined IoT | |
| Fang et al. | Latency aware online tasks scheduling policy for edge computing system | |
| CN115190082A (en) | Traffic flow scheduling method and device for TSN (traffic service network) | |
| CN110086662B (en) | Method for implementing demand definition network and network architecture |
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 |