CN113965568B - Edge computing system for urban road C-V2X network - Google Patents
Edge computing system for urban road C-V2X network Download PDFInfo
- Publication number
- CN113965568B CN113965568B CN202111213957.0A CN202111213957A CN113965568B CN 113965568 B CN113965568 B CN 113965568B CN 202111213957 A CN202111213957 A CN 202111213957A CN 113965568 B CN113965568 B CN 113965568B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- edge computing
- queue
- traffic
- edge
- 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
- 238000012545 processing Methods 0.000 claims abstract description 35
- 238000001514 detection method Methods 0.000 claims abstract description 22
- 238000013468 resource allocation Methods 0.000 claims abstract description 22
- 238000003860 storage Methods 0.000 claims abstract description 19
- 230000004927 fusion Effects 0.000 claims abstract description 14
- 238000004891 communication Methods 0.000 claims abstract description 11
- 230000008447 perception Effects 0.000 claims abstract description 10
- 238000004458 analytical method Methods 0.000 claims abstract description 9
- 238000004364 calculation method Methods 0.000 claims description 63
- 238000000034 method Methods 0.000 claims description 21
- 238000007726 management method Methods 0.000 claims description 19
- 230000006855 networking Effects 0.000 claims description 14
- 230000001413 cellular effect Effects 0.000 claims description 12
- 230000005540 biological transmission Effects 0.000 claims description 10
- 238000005516 engineering process Methods 0.000 claims description 9
- 230000033001 locomotion Effects 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000005457 optimization Methods 0.000 claims description 9
- 238000011156 evaluation Methods 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 7
- 238000012544 monitoring process Methods 0.000 claims description 6
- 239000013307 optical fiber Substances 0.000 claims description 6
- 230000003068 static effect Effects 0.000 claims description 6
- 239000013598 vector Substances 0.000 claims description 6
- 230000001133 acceleration Effects 0.000 claims description 5
- 238000002955 isolation Methods 0.000 claims description 4
- 230000009471 action Effects 0.000 claims description 3
- 238000010191 image analysis Methods 0.000 claims description 3
- 238000013508 migration Methods 0.000 claims description 3
- 230000005012 migration Effects 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000007405 data analysis Methods 0.000 claims description 2
- 230000007613 environmental effect Effects 0.000 claims description 2
- 230000010354 integration Effects 0.000 claims 1
- 230000006870 function Effects 0.000 abstract description 2
- 230000002349 favourable effect Effects 0.000 abstract 1
- 238000013473 artificial intelligence Methods 0.000 description 11
- 230000008901 benefit Effects 0.000 description 6
- 238000013461 design Methods 0.000 description 5
- 238000013135 deep learning Methods 0.000 description 4
- 230000006872 improvement Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000010267 cellular communication Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000006698 induction Effects 0.000 description 2
- 210000001503 joint Anatomy 0.000 description 2
- 229920003087 methylethyl cellulose Polymers 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001934 delay Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000003032 molecular docking Methods 0.000 description 1
- 238000010223 real-time analysis Methods 0.000 description 1
- 229920006395 saturated elastomer Polymers 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
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
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0803—Configuration setting
- H04L41/0823—Configuration setting characterised by the purposes of a change of settings, e.g. optimising configuration for enhancing reliability
- H04L41/0826—Configuration setting characterised by the purposes of a change of settings, e.g. optimising configuration for enhancing reliability for reduction of network costs
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/30—Services specially adapted for particular environments, situations or purposes
- H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
- H04W4/44—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P] for communication between vehicles and infrastructures, e.g. vehicle-to-cloud [V2C] or vehicle-to-home [V2H]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/30—Services specially adapted for particular environments, situations or purposes
- H04W4/40—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P]
- H04W4/46—Services specially adapted for particular environments, situations or purposes for vehicles, e.g. vehicle-to-pedestrians [V2P] for vehicle-to-vehicle communication [V2V]
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses an edge computing system for an urban road C-V2X network, which comprises three layers of a vehicle end MEC unit, a road side MEC unit and a base station MEC unit and a distributed resource allocation and scheduling strategy, wherein the three layers cooperate to realize five types of edge computing functions of traffic target detection and identification, traffic event discovery and prediction, traffic scene fusion analysis and processing, local traffic situation comprehensive perception, auxiliary and automatic driving decision for the urban road C-V2X network. The edge computing system can clearly divide and accurately meet the functional requirements of C-V2X network operation on the edge computing unit in urban road traffic environment, is favorable for reasonable distribution and scheduling of computing power, storage, network and application resources of edge computing equipment on vehicles, roads and side ends, furthest plays the capability of edge computing in improving the application safety, efficiency and service quality of the road traffic intelligent network, and lays a foundation for realizing perception, computation and communication integrated vehicle-road cooperative systems in the future.
Description
Technical Field
The invention relates to the fields of edge calculation, internet of vehicles and vehicle-road coordination, in particular to an edge calculation system for an urban road C-V2X network.
Background
Worldwide, urban road traffic is being saturated with the trouble of traffic jams, frequent accidents, and other aeipatory diseases. The maintenance of motor vehicles has been growing rapidly year by year, and the mere increase of road supply to alleviate traffic problems has proven to be limiting. Intelligent traffic comprehensively utilizes modern communication, perception, calculation, network exchange, new energy automobiles, automatic driving, big data and other technologies, can be used as an effective supplement for solving the difficult problem, vehicles become more and more intelligent by means of the high and new technologies, and roads become more and more intelligent.
Edge computing has received a great deal of attention in recent years in both academia and industry, with distributed transformation of computing power and resources deployed close to customers having been a trend. The edge calculation has incomparable advantages in the aspects of saving data bandwidth, reducing backhaul capacity, reducing transmission cost, reducing time delay, improving data real-time analysis capability, protecting user data privacy and the like, so that the edge calculation can provide stronger calculation capability and more optimized analysis processing means for intelligent transportation together with cloud calculation.
Due to the diversity of application scenarios, there is no unified solution for a specific implementation architecture of edge computation. The main influencing factors include specific requirements (time delay, bandwidth, real-time performance, data transmission, security), technical conditions (edge configuration, distance from cloud and terminal equipment), service characteristics (requirements, economic considerations) and the like of the application. For example, only for the location of the "edge" is understood, there are terminal devices, service sites, nearby base stations, aggregation points, transport networks, core networks, near cloud, etc. Communication equipment providers, IT manufacturers, operators, factories, solution providers and the like all promote different edge computing systems according to own industry advantages, and the systems have various characteristics, but in terms of understanding and targeted design of traffic service, a set of intelligent traffic special edge computing systems with clear hierarchical design, reasonable functional division and strong scene directivity is also lacking.
Disclosure of Invention
The invention aims to: aiming at the defects of the prior art, the invention provides an edge computing system for an urban road C-V2X (Cellular-Vehicle to Everything) network.
In order to solve the technical problems, the invention discloses an edge computing system for an urban road C-V2X network, which comprises an infrastructure layer, a network layer, a resource layer, a platform layer, a scene layer and a service application layer, wherein the infrastructure layer comprises three layers of a vehicle end MEC (Multi-Access Edge Computing ) unit, a road side MEC unit and a base station MEC unit and a set of distributed resource allocation and scheduling strategy, and all MEC units are connected through a C-V2X network and a wired network to form a vehicle, road and station integrated deployment architecture; edge computing tasks are processed among MEC units through a distributed resource allocation and scheduling strategy.
The vehicle-end MEC Unit comprises a special vehicle-mounted MEC terminal, a vehicle-mounted computer and a vehicle-mounted OBU (On board Unit) which are deployed at the vehicle end and is used for processing an edge calculation task approaching the vehicle target On a road;
the road side MEC unit comprises portable computers which are arranged on road side signal lamp poles, monitoring poles, special stand columns and a chassis, is generally of an embedded architecture, can support AI deep learning and the like, and is used for processing edge calculation tasks of more than two traffic targets in road openings and road sections from the road side;
the base station MEC unit comprises a general edge calculation server which is deployed in a cellular network base station room and is used for processing edge calculation tasks of a large number of traffic targets related to situation awareness and cooperative control in the coverage area of the base station;
the edge calculation tasks comprise traffic target detection and identification, traffic event discovery and prediction, traffic scene fusion analysis and processing, local traffic situation comprehensive perception and auxiliary and automatic driving decision tasks on task categories.
The distributed resource allocation and scheduling policy refers to a policy for distributed configuration and management of edge computing resources (including computing power, bandwidth, storage, etc.) in a C-V2X network. The general principle is to flexibly adopt technologies such as virtual machines, containers, micro-services, heterogeneous computing, deep learning and the like, comprehensively consider factors such as geographic positions, network time delay, message periods, event properties, complexity, transmission conditions and the like, and dynamically allocate and schedule resources according to specific scene demands. The same set of general distributed resource allocation and scheduling strategies are preset in various MECs to support flexible and rapid scene design development and deployment operation, and effectively support scene edge calculation tasks such as traffic target detection and identification, traffic event discovery and prediction, intelligent network business scene realization, local traffic situation comprehensive perception, assistance, automatic driving decision and the like.
The three-level edge computing units are subjected to scientific, reasonable, flexible and efficient distributed management, so that five types of traffic edge computing tasks of traffic target detection and identification, traffic event discovery and prediction, traffic scene fusion analysis and processing, local traffic situation comprehensive perception, assistance and automatic driving decision are respectively realized.
In one implementation, a dedicated on-board MEC terminal includes a lightweight edge computing device dedicated to vehicle sensory data analysis processing and simple fusion calculations. The method is characterized in that abundant external interfaces are required to be provided for docking with devices such as a vehicle-mounted camera, a millimeter wave radar, a laser radar, an ultrasonic radar, a CAN (controller area network) bus, a Beidou/GPS (Global Positioning System ) positioning device, a vehicle-mounted OBU (on-board) and the like, the calculation power supports the fusion processing of the original data perceived by the docked vehicle-mounted camera, the millimeter wave radar, the laser radar, the ultrasonic radar, the CAN bus and the Beidou/GPS positioning device, real-time traffic target detection and identification are completed, dynamic and static characteristics such as the distance, the speed, the direction, the acceleration, the outline, the color, the license plate number and the like of the target are included, the dynamic and static characteristics are transmitted to the vehicle-mounted OBU in a structured data form, and the data are transmitted outwards by the OBU.
The vehicle-mounted computer comprises a highly integrated vehicle-gauge-level special-purpose computer, and meets strict requirements on temperature environment, vibration and impact resistance, reliability, consistency, manufacturing process and the like. The computer is installed in front of a vehicle factory, and software and hardware are relatively closed, so that the computer is mainly used for carrying out real-time dynamic monitoring, fault detection and early warning on the working state of the vehicle.
The vehicle-mounted OBU comprises vehicle-mounted communication equipment special for the vehicle networking, which adopts the C-V2X technology and has certain marginal computing power. The edge computing force can be matched with a special vehicle-mounted MEC terminal to cooperatively complete a large number of complex vehicle-end computing tasks. The vehicle-mounted OBU can interact data of the vehicle-end MEC unit with the base station MEC unit through the cellular network, and interact with other vehicle-end MEC units and road side MEC units through the V2X network, so that efficient coordination among the MEC units is realized. The cellular network refers to a 4G/5G cellular communication network, namely a Uu cellular communication mode of a C-V2X network; the V2X network refers to a PC5 direct communication mode of the C-V2X network.
In one implementation, the roadside MEC unit requires industrial-scale design to ensure its operational reliability and stability due to the harsh operating environment requirements. The Road Side MEC Unit is mainly in butt joint with an RSU (Road Side Unit), a Road Side camera, a millimeter wave radar, a laser radar, an information board/induction screen, a Beidou/GPS, a annunciator, a weather sensor, an environment sensor and the like through an Ethernet switch, and is connected with the base station MEC through an optical fiber or a cellular network.
The base station MEC units may be deployed on a rack or separately depending on the particular equipment and site conditions. And in the base station where the MEC equipment is deployed by the operator, the MEC unit can select to multiplex the operator equipment and share the resources of calculation power, storage, network and the like so as to save the cost and improve the benefit. The base station MEC may be connected to the roadside MEC unit via an optical fiber or cellular network, and to the vehicle end MEC via a cellular network.
In one implementation, the resource layer includes a heterogeneous resource sub-layer, a resource allocation sub-layer and a resource scheduling sub-layer, the heterogeneous resource sub-layer includes a heterogeneous processor involved in edge computation, a unified representation of network bandwidth and storage resources, and the resource allocation sub-layer performs virtualization configuration management, including a virtual machine mode and a containerization mode; the resource scheduling sub-layer performs resource calculation, resource scheduling, resource isolation, scheduling optimization, job queue management, load balancing, virtual machine migration and resource unloading according to the edge calculation task characteristics;
the platform layer deploys micro-services according to the requirements of real-time performance, security and heterogeneous computation of each MEC unit, and the platform layer supporting component comprises service registration, service discovery, service gateway, service arrangement, API (Application Programming Interface, application program interface) management, an integrated framework, distributed management and a call chain; the platform layer also provides traffic AI (Artificial Intelligence ) algorithm support and security authentication;
The scene layer provides calculation support of intelligent networking edge scenes including traffic safety, traffic efficiency, travel service and automatic driving, and can be elastically expanded according to requirements;
the service application layer is cooperated with the cloud service application layer on the basis of scene layer service, and provides intelligent networking and intelligent traffic service applications including bus priority, holographic intersections, traffic sequencing, digital twinning, signal optimization and free flow charging for traffic managers, traffic transportation industries and driver users.
In one implementation, the distributed resource allocation and scheduling policy includes:
step 1, determining business application, and defining the related traffic participation object and the action range of the business application;
dividing the edge scenes, and dividing the determined business application into different edge scene application combinations;
step 3, analyzing detailed characteristics of the edge scene application, wherein the detailed characteristics comprise application properties, event types, high-precision map and positioning requirements, time delay requirements, reliability requirements, transmission bandwidth requirements, message periods, data packet sizes, data types, AI requirements, multi-source fusion requirements, decision and control requirements, road characteristics, target motion characteristics, static characteristics, environment and weather influences of the edge scene application;
Step 4, determining an edge computing task, and determining the edge computing task for realizing the edge scene according to the detailed characteristics of the edge scene application;
step 5, determining the detailed description of the edge computing resources, submitting an edge computing task request to a platform layer, calling a corresponding micro-service and algorithm model by the platform layer through a micro-service API interface gateway according to the request, and issuing the detailed description of the required computing resources to a resource layer through the micro-service;
and 6, carrying out distributed resource scheduling, carrying out fine scheduling of computing power, bandwidth and storage according to the detailed description of the required edge computing resources by a resource layer, and distributing the proper resources to the proper edge computing tasks.
In one implementation, the edge computing tasks include latency-sensitive, security-sensitive, bandwidth-sensitive, power-consuming, memory-footprint, and heterogeneous collaborative tasks in terms of task characteristics.
In one implementation, distributed resource scheduling is performed, a scheduling scheme based on a distributed resource scheduling algorithm and a combination thereof are preset aiming at different task characteristics of edge computing tasks by integrating the existing resource pool conditions, a menu type resource scheduling scheme service is provided, trivial resource scheduling details are shielded for upper-layer application, and scheduling acceleration and optimization are realized; for a vehicle-end MEC unit, computing resources of the vehicle-end MEC unit are preferentially distributed to the vehicle; for a road side MEC unit, the supported resource scheduling range is in the signal coverage range of the RSU, and is used for scene calculation including road side image analysis, laser radar point cloud processing algorithm and high-precision map matching, and supporting the provision of calculation resources from the road side to the vehicle end, so as to realize the cooperative auxiliary driving and automatic driving support of the vehicle and road; for the base station MEC unit, scene calculations for long periods (minutes), medium delays (20 ms-200 ms) and weak locality (across multiple adjacent intersections).
In one implementation, the distributed resource scheduling algorithm includes:
step 6.1, initializing a queue, dividing an edge computing task into 6 queues according to task characteristics, and recording the queue where a delay sensitive task is located as r 1 The queue where the security sensitive task is located is r 2 The queue where the bandwidth sensitive task is located is r 3 The queue where the task with power consumption is located is r 4 The queue where the memory occupied task is located is r 5 And the queue where the heterogeneous collaborative task is located is r 6 Each queue is initialized to allocate a fixed lengthA degree; edge computing task requests with different task characteristics enter corresponding queues in parallel;
step 6.2, setting the priority of each queue, allocating resources to the queues with high authority preferentially, and defining the priority of each queue as r in sequence 1 >r 2 >r 3 ＝r 4 ＝r 5 ＝r 6 I.e. system priority handling queue r 1 Is a request in (a); in the process of receiving the request, if the queue with the highest priority reaches the allocation length, combining the queue with the next-priority queue; if a certain queue is empty in a set time interval, merging the queue with the highest priority; after the request in the highest priority queue is completely processed, processing the next priority queue;
step 6.3, determining the scheduling sequence of different requests in the same queue;
And 6.4, carrying out resource allocation processing on the requests in each queue according to the scheduling sequence.
In one implementation, the step 6.3 includes:
step 6.3.1, defining QoS model indexes of the edge computing system, wherein the model indexes comprise computing time, transmission time, scheduling time, bandwidth overhead, computing power resources, storage resources, security requirements and authority indexes;
step 6.3.2, defining vector s= { S 1 ,s 2 ,...,s i ,...,s n The request service set in a certain queue is shown, n is the number of requests, i is more than or equal to 1 and less than or equal to n; vector q= { Q 1 ,q 2 ,...,q j ,...,q m The number m is the total number of QoS model indexes, and j is more than or equal to 1 and less than or equal to m; the weight matrix is p= (P ij ) n×m ，p ij Representing the importance requirement of the ith request in the queue on the jth QoS index;
step 6.3.3, normalizing the weight matrix to obtain a calculation matrix Y= (Y) ij ) n×m ：
Wherein,,
maxp .j ＝max{p ij },1≤i≤n，minp .j ＝min{p ij },1≤i≤n；
step 6.3.4, calculating the information entropy value H of the ith request i The method comprises the following steps:
step 6.3.5, calculating an evaluation index w of the ith request i The method comprises the following steps:
step 6.3.6, requests to each queue are according to the evaluation index w i And (5) arranging in a descending order to obtain the scheduling orders of different requests.
In one implementation, after the distributed resource scheduling is completed in step 6, that is, after the edge computing task is completed, the corresponding container and its mirror image are deleted, and the occupied resources thereof are released in time.
The beneficial effects are that:
according to the invention, by integrating three different types and levels of edge computing equipment of the vehicle-end MEC unit, the road-side MEC unit and the base station MEC unit, the scientific and reasonable configuration of the edge computing resources at the vehicle-end, the road-side and the base station is realized by utilizing the distributed resource allocation and scheduling strategy designed for traffic application scenes, the dynamic utilization efficiency of the resources is improved, and the overall deployment cost of the edge computing is reduced. The edge computing tasks are divided into different categories such as a target detection and identification category, an event discovery and prediction category, a service scene implementation category, a comprehensive situation awareness category, an auxiliary and automatic driving decision category and the like, so that the resource allocation and scheduling strategies are determined in a targeted manner, and the traffic edge computing deployment targets of 'tasks according to scenes', 'strategies according to tasks', 'resources according to strategies' are realized.
The invention can provide a multi-type hierarchical edge computing architecture which is close to the actual traffic scene requirement for the urban road C-V2X network by utilizing the technologies of edge computing, distributed resource allocation and scheduling, virtual machines, containerization, micro-service, heterogeneous computing, deep learning and the like, can effectively improve the edge computing efficiency and the edge resource utilization rate, reduce scene analysis time delay, improve the reliability of cooperative application of the roads, simultaneously reduce the overall deployment cost and improve the project fund use benefit.
Drawings
The foregoing and/or other advantages of the invention will become more apparent from the following detailed description of the invention when taken in conjunction with the accompanying drawings and detailed description.
FIG. 1 is a schematic diagram of a physical system architecture of the present invention.
Fig. 2 is a schematic diagram of the overall technical architecture of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.
The embodiment of the application provides an edge computing system for an urban road C-V2X network, which is shown in fig. 2 and comprises an infrastructure layer, a network layer, a resource layer, a platform layer, a scene layer and a service application layer, wherein the infrastructure layer comprises a vehicle-end MEC unit, a road-side MEC unit and a base station MEC unit, and the MEC units are tightly connected through the C-V2X network, a wired (Ethernet/optical fiber) network and the like to form a vehicle-road-station integrated deployment architecture as shown in fig. 1. Through a distributed resource allocation and scheduling strategy, each MEC unit orderly organizes resources, operates algorithms, programs business, interactive communication, stores data and the like, and achieves various types of edge computing tasks such as target detection and identification class, event discovery and prediction class, business scene implementation class, comprehensive situation awareness class, auxiliary and automatic driving decision class and the like.
The vehicle-end MEC unit comprises a special vehicle-mounted MEC terminal, a vehicle-mounted computer and a vehicle-mounted OBU which are deployed at the vehicle end and is used for processing an edge calculation task approaching the vehicle target on a road;
the road side MEC unit comprises a portable computer which is arranged on a road side signal lamp post, a monitoring rod, a special stand column and a chassis and is used for processing edge calculation tasks of more than two traffic targets in a road side opposite port and a road section;
the base station MEC unit comprises a general edge calculation server which is deployed in a cellular network base station room and is used for processing edge calculation tasks of a large number of traffic targets related to situation awareness and cooperative control in the coverage area of the base station;
the distributed resource allocation and scheduling strategy refers to a strategy for performing distributed configuration and management on all edge computing resources in a C-V2X network and a wired network, wherein the edge computing resources comprise computing power, bandwidth and storage;
the vehicle-end MEC unit focuses on processing tasks such as identification detection of approaching vehicle targets, safety event early warning, emergency decision, automatic driving and the like on a road, and the road-side MEC unit focuses on completing tasks such as identification detection of multiple traffic targets in intersections and road sections from the road side, multi-element perception data fusion, event early warning, blind area and beyond-sight risk reminding, safety and efficiency improvement and auxiliary driving scene implementation. The base station MEC unit focuses on completing tasks of controlling overall traffic situation in coverage, fusing multi-source event information, realizing intelligent networking service (such as bus priority, holographic intersection and the like), issuing and updating high-precision map, issuing static information, issuing quasi-static information and the like. It should be noted that the tasks of the above types may have certain overlapping intersections in the MECs of the three layers, but the view angles, the departure points and the action ranges are all focused, and cannot be replaced with each other.
In this embodiment, the special vehicle-mounted MEC terminal includes a lightweight edge computing device specifically used for analysis processing and simple fusion computation of vehicle sensing data, where the lightweight edge computing device is provided with an external interface, and is in butt joint with a vehicle-mounted camera, a millimeter wave radar, a laser radar, an ultrasonic radar, a CAN bus, a beidou/GPS positioning device and a vehicle-mounted OBU device through the external interface; the computing power support of the lightweight edge computing equipment fuses the original data perceived by the butted vehicle-mounted camera, the millimeter wave radar, the laser radar, the ultrasonic radar, the CAN bus and the Beidou/GPS positioning equipment to finish real-time traffic target detection and identification, wherein the real-time traffic target detection and identification comprises the characteristics of the distance, the speed, the direction, the acceleration, the outline, the color and the license plate number of a target, and the real-time traffic target detection and identification are transmitted to the vehicle-mounted OBU in a structured data form;
the vehicle-mounted computer comprises a vehicle-mounted special computer, wherein the vehicle-mounted special computer is installed in front of a vehicle factory, and the software and the hardware are relatively closed and are used for carrying out real-time dynamic monitoring and fault detection and early warning on the working state of the vehicle;
the vehicle-mounted OBU comprises vehicle-mounted communication equipment special for the vehicle networking, which adopts a C-V2X technology and has edge calculation capability, and the edge calculation capability of the vehicle-mounted communication equipment special for the vehicle networking can be matched with a special vehicle-mounted MEC terminal to cooperatively complete a vehicle-end edge calculation task; as shown in fig. 1, the on-board OBU can interact data of a vehicle-end MEC unit with a base station MEC unit through a cellular network, and interact with other vehicle-end MEC units and road-side MEC units through a V2X network.
In this embodiment, as shown in fig. 1, the roadside MEC unit is docked with the RSU, the roadside camera, the millimeter wave radar, the laser radar, the information board/the induction screen, the beidou/GPS, the annunciator, the weather sensor and the environmental sensor through the ethernet switch, and is connected with the base station MEC unit through the optical fiber or the cellular network.
As shown in fig. 2, the overall technical architecture of the present embodiment mainly comprises an infrastructure layer, a network layer, a resource layer, a platform layer, a scene layer, and a service application layer. The infrastructure layer mainly comprises a vehicle end MEC, a road side MEC, a base station MEC, and equipment such as perception and communication accessed by the base station MEC, wherein the network layer comprises three networking modes of C-V2X Uu, C-V2X PC5 and a wired network (Ethernet/optical fiber), the resource layer comprises a heterogeneous resource sub-layer, a resource configuration sub-layer and a resource scheduling sub-layer, the heterogeneous resource sub-layer mainly comprises various types of heterogeneous processors possibly related to edge calculation, unified representation of network bandwidth and storage resources, and the resource configuration sub-layer performs virtualization configuration management, including a virtual machine mode and a containerization mode. The virtual machine mode includes virtual machine software VMware (Virtual Machine ware), an open source virtual machine KVM (Kernel-Based Virtual Machine), a virtual box and an open source cloud computing management platform project OpenStack, and the containerization mode includes an application container engine Docker and container cluster management/orchestration tools K8S (Kubernetes), compose, marathon and Swarm. The lightweight nature of edge computing, relative to cloud computing, primarily uses container technology to support the running environment of micro-services, but in combination with virtual machine technology can provide more secure isolation for containers. The resource scheduling sub-layer performs resource calculation, resource scheduling, resource isolation, scheduling optimization, job queue management, load balancing, virtual machine migration, resource unloading and the like according to the task characteristics of edge calculation. The platform layer is mainly used for deploying a large amount of micro-services according to the requirements on the real-time performance, the security and the heterogeneous computation of the MEC units, and the support components of the micro-services comprise service registration, service discovery, service gateway, service arrangement, API management, an integrated framework, distributed management, a call chain and the like. The platform also needs to provide traffic AI algorithm support, security authentication (including software-based security authentication and underlying hardware-based security trust root, etc.), and so forth. The scene layer provides calculation support for tens of traffic safety, traffic efficiency, travel service and automatic driving intelligent networking scenes, and can be elastically expanded according to requirements. The business application layer cooperates with the cloud business application layer on the basis of scene layer service, and provides intelligent networking and intelligent traffic business applications including bus priority, holographic intersections, traffic sequencing, digital twinning, signal optimization and free flow charging for users such as traffic managers, traffic transportation industries, drivers and the like.
The distributed resource allocation and scheduling strategy of the invention is as follows:
step 1, determining specific business application
Specific business applications to be realized, such as bus priority, holographic intersections, signal optimization and the like, are determined according to the design function; the traffic participation objects explicitly involved include departments, personnel, vehicles, facilities, equipment, terminals and the like; it should also be clear that the scope of service applications, such as single points, trunks, intersections, road segments, areas, cities, etc., is limited to the coverage of a single base station for the architecture of the present invention.
Step 2, partitioning supported edge scene applications
According to the definition and meaning of the business application, the business application is divided into different edge scene application combinations in detail, and the scene application is generally a standardized scene supported by the existing standard framework and can be cut and expanded according to the actual business requirements.
Step 3, analyzing detailed characteristics of the edge scene application
For the divided scene application, the detailed characteristics of the scene application are further analyzed, including application properties (safety, efficiency, service and automatic driving), event types, high-precision map and positioning requirements, time delay requirements, reliability requirements, transmission bandwidth requirements, message periods, data packet sizes, data types (whether structured data and streaming media data) and the like, AI requirements, multi-source fusion requirements, decision and control requirements, road characteristics, target motion characteristics, static characteristics, environment and weather influences and the like.
Step 4, determining specific requirements of scene edge computing tasks
And determining the specific edge computing task requirements needed for realizing the scene according to the characteristics of the scene application. The method generally comprises five types of scene calculation demands of traffic target detection and identification, traffic event discovery and prediction, traffic scene fusion analysis and processing, local traffic situation comprehensive perception, assistance and automatic driving decision; the method is characterized by being divided into time delay sensitive type, safety sensitive type, bandwidth sensitive type, calculation power consumption type, storage occupation type, heterogeneous cooperation type and other task demands. For example, intersection anti-collision, emergency brake early warning and the like are typical time delay sensitive task requirements, illegal behavior evidence collection and high-precision map matching are typical storage occupation type task requirements, and video AI analysis is a typical computational power consumption task requirement.
Step 5, determining a detailed description of the required edge computing resources
After the task requirement is determined, an edge computing task request can be submitted to the platform layer, and the platform layer calls a corresponding micro-service and algorithm model through a micro-service API interface gateway according to the request and issues detailed description of required computing resources to the resource layer through the micro-service.
Step 6, carrying out distributed resource scheduling, and distributing proper edge computing resources to proper tasks in a micro-service mode
The resource layer performs fine scheduling of computing power, bandwidth and storage according to the detailed description of the required edge computing resources, and allocates the appropriate resources to the appropriate edge computing tasks. The process is to pre-formulate a series of scheduling schemes and combinations thereof based on a distributed resource scheduling algorithm by integrating the existing resource pool conditions and calculating task characteristics aiming at typical traffic edges, provide a menu-type resource scheduling scheme service, shield trivial and variable resource scheduling details for upper-layer application, and realize the acceleration and optimization of scheduling. The scheduling scheme set is pre-configured on each distributed edge computing node, has the characteristics of dynamic property, flexibility, scene directivity, resource correlation and the like, and is the core of a scheduling strategy. In general, a CPU (Central Processing Unit ) is suitable for tasks such as decision control and rule management, a GPU (Graphics Processing Unit, graphics processor) is suitable for tasks requiring massive parallel computation such as AI training and target detection, a DSP (Digital Signal Processing ) is suitable for image and video processing, an FPGA (Field Programmable Gate Array ) is suitable for tasks such as multi-source information fusion, video segmentation, target tracking and event prediction, and has programmable and easily-updated characteristics, and an ASIC (Application Specific Integrated Circuit ) is suitable for custom vision processing and AI algorithm tasks with powerful performance and mature stability. Such as NPU (Neural-network Processing Unit, neural network processing unit) and TPU (Tensor Processing Unit ), belong to ASIC processors for AI deep learning. For the vehicle end MEC, because it is in a high-speed motion state and the allocated mobile bandwidth resources are limited, in principle, the computing resources are preferentially allocated to the vehicle. If there is an intelligent network-connected vehicle supporting the architecture near the intelligent network-connected vehicle, the edge resources can support sharing, but the network rapid switching caused by the movement of the vehicle needs to be considered, and the resources should support rapid allocation and unloading. For road side MEC, the supported resource scheduling range is generally in the signal coverage range of RSU, and is generally used for scene calculation such as road side image analysis, laser radar point cloud processing algorithm, high-precision map matching and the like, and supporting the provision of calculation resources from the road side to the vehicle end, so that the cooperative auxiliary driving and automatic driving support of the vehicle and the vehicle end calculation force are less utilized. In addition, the road side MEC and the base station MEC have two networking modes of wireless and wired, the position is fixed, the sharing of calculation power, bandwidth and storage resources is easy to realize, and more flexible resource scheduling can be realized. For example, the scene calculation with short period, low time delay and strong locality can be completed at the road side MEC, while the scene calculation with long period, medium time delay and weak locality can be completed at the base station MEC. For another example, training and modeling processes in the AI scenario may be done at the base station MEC, while real-time inference processes are done at the roadside MEC. The base station MEC is usually a server or virtualized edge cloud of an x86 architecture, the architecture is relatively single and rich in resources, and the base station MEC can play a role of a certain 'resource buffer zone' and a certain 'regional field dispatching center' and a certain 'security authentication center' in the architecture. The above scheduling process is transparent to the end user due to the nature of the resource virtualization.
Taking an intersection anti-collision scene as an example, the edge calculation tasks related to the scene generally comprise main vehicle motion state sensing, motor vehicles, non-motor vehicles on an intersection road, pedestrian target detection and identification, motion parameter extraction, track prediction, collision early warning, event reporting and storage and the like. Most tasks have high latency and security requirements. At the moment, the main vehicle motion state sensing is suitable for directly acquiring driving computer data by the vehicle end OBU and sending the driving computer data to the RSU, the motor vehicles, non-motor vehicles, pedestrian targets on the intersecting road are large in calculation power consumption such as motion parameter extraction, track prediction and collision early warning, the time delay requirement is high, the road side MEC is suitable for carrying out fusion sensing on data such as road side video and millimeter wave radar, map matching data and data reported by the vehicle end OBU, the result is broadcasted to the road side vehicle, and the main vehicle sends early warning to a driver after receiving the broadcast. When a collision event cannot be avoided, event original data needs to be uploaded to a cloud control platform, the task belongs to a bandwidth sensitive type task and a storage occupied type task, but the time delay requirement can be relaxed, and the calculation priority is lower. In addition, the dynamic update of the map data belongs to the tasks of time delay insensitivity, long period and burst transmission, and can be issued to the road side MEC at fixed time by the cloud control platform or the base station MEC.
The edge calculation task in the urban road C-V2X network has the characteristics of strong real-time performance, high concurrency and multiple sexualization, has higher requirements on the waiting time, service quality, load balancing, resource utilization rate and the like of edge calculation, and cannot meet the requirements by the traditional resource scheduling algorithm.
The invention adopts a weighting method to carry out optimized scheduling on resources, combines the scheduling characteristics of the vehicle and the road under different scenes, and uses a selection model to convert the scheduling problem into a multi-attribute decision problem so as to realize real-time dynamic scheduling.
Aiming at the scheduling strategy, a distributed resource scheduling algorithm is designed as follows:
step 6.1, initializing a queue, dividing an edge computing task into 6 queues according to task characteristics, and recording the queue where a delay sensitive task is located as r 1 The queue where the security sensitive task is located is r 2 The queue where the bandwidth sensitive task is located is r 3 The queue where the task with power consumption is located is r 4 The queue where the memory occupied task is located is r 5 And the queue where the heterogeneous collaborative task is located is r 6 Each queue is initialized, a fixed length (total number of requests in the queue) n is allocated, a large part of message request frequencies in the Internet of vehicles are fixed, and n is determined according to known fixed message request frequencies and estimated random message request frequencies in combination with scene requirements; different types of requests enter different queues in parallel.
And 6.2, presetting the priority of each queue, and preferentially distributing resources to the queues with high authority. According to the resource demand characteristics of the vehicle-road cooperative tasks, sequentially defining each type of priority as r 1 >r 2 >r 3 ＝r 4 ＝r 5 ＝r 6 I.e. system priority handling queue r 1 Is a request for a request in (a). In the process of receiving the request, if the queue with the highest priority reaches the allocation length, combining the queue with the next-priority queue, namely, jointly using the length of 2n for the requests in the queues r1 and r 2; if a certain queue is empty in a set time interval, merging the queue with the highest priority queue, namely temporarily canceling the priority in the set time interval, and increasing the length of the request queue with the highest priority; after the request in the highest priority queue is completely processed, the next priority queue is processed. In this way, the success rate of high priority requests is improved.
Step 6.3, determining the scheduling sequence of different requests in the same queue;
step 6.3.1, defining typical QoS model indexes of the edge computing system, wherein the model indexes comprise computing time, transmission time, scheduling time, bandwidth overhead, computing power resources, storage resources, security requirements and authority indexes;
step 6.3.2, defining vector s= { S 1 ,s 2 ,...,s i ,...,s n The request service set in a certain queue is shown, n is the number of requests, i is more than or equal to 1 and less than or equal to n; vector q= { Q 1 ,q 2 ,...,q j ,...,q m The number m is the total number of QoS model indexes, and j is more than or equal to 1 and less than or equal to m; the weight matrix is p= (P ij ) n×m ，p ij The importance requirement of the ith request in the queue on the jth QoS index is expressed, and the weight can be determined by an expert evaluation method and can be adjusted in application according to the actual requirements of a scene.
Step 6.3.3, normalizing the weight matrix to obtain a calculation matrix Y= (Y) ij ) n×m ：
Wherein,,
maxp .j ＝max{p ij },1≤i≤n，minp .j ＝min{p ij },1≤i≤n；
step 6.3.4, calculating the information entropy value H of the ith request i The method comprises the following steps:
step 6.3.5, calculating an evaluation index w of the ith request i The method comprises the following steps:
step 6.3.6, requests to each queue are according to the evaluation index w i And (5) arranging in a descending order to obtain the scheduling orders of different requests.
And 6.4, carrying out resource allocation processing on the requests in each queue according to the scheduling sequence.
And 7, completing the task and releasing the resources. After the edge computing task is completed, deleting the corresponding container and the mirror image thereof, and timely releasing the occupied resources.
The invention provides an edge computing system based on an urban road C-V2X network, and a plurality of methods and approaches for realizing the architecture are provided, the method is just one of the embodiments of the invention, and it should be noted that, for a person skilled in the art, a plurality of improvements and modifications can be made without departing from the principle of the invention, and the improvements and modifications should also be regarded as the protection scope of the invention. The components not explicitly described in this embodiment can be realized by the prior art.
Claims (8)
1. An edge computing system for an urban road C-V2X network comprises an infrastructure layer, a network layer, a resource layer, a platform layer, a scene layer and a service application layer, and is characterized in that the infrastructure layer comprises a vehicle-end MEC unit, a road-side MEC unit and a base station MEC unit, and all MEC units are connected with a wired network through the C-V2X network to form a vehicle, road and station integrated deployment architecture; processing edge computing tasks among MEC units through a distributed resource allocation and scheduling strategy;
the vehicle-end MEC unit comprises a special vehicle-mounted MEC terminal, a vehicle-mounted computer and a vehicle-mounted OBU which are deployed at the vehicle end and is used for processing an edge calculation task approaching the vehicle target on a road;
the road side MEC unit comprises a portable computer which is arranged on a road side signal lamp post, a monitoring rod, a special stand column and a chassis and is used for processing edge calculation tasks of more than two traffic targets in a road side opposite port and a road section;
the base station MEC unit comprises a general edge calculation server which is deployed in a cellular network base station room and is used for processing edge calculation tasks of a large number of traffic targets related to situation awareness and cooperative control in the coverage area of the base station;
the distributed resource allocation and scheduling strategy refers to a strategy for performing distributed configuration and management on all edge computing resources in a C-V2X network and a wired network, wherein the edge computing resources comprise computing power, bandwidth and storage;
The edge calculation task comprises traffic target detection and identification, traffic event discovery and prediction, traffic scene fusion analysis and processing, local traffic situation comprehensive perception and auxiliary and automatic driving decision tasks on task categories;
the resource layer comprises a heterogeneous resource sub-layer, a resource allocation sub-layer and a resource scheduling sub-layer, wherein the heterogeneous resource sub-layer comprises a heterogeneous processor, a network bandwidth and a unified representation of storage resources, which are involved in edge calculation, and the resource allocation sub-layer performs virtualization configuration management, including a virtual machine mode and a containerization mode; the resource scheduling sub-layer performs resource calculation, resource scheduling, resource isolation, scheduling optimization, job queue management, load balancing, virtual machine migration and resource unloading according to the edge calculation task characteristics;
the platform layer deploys micro-services according to the requirements of real-time performance, security and heterogeneous computation of each MEC unit, and the platform layer supporting component comprises service registration, service discovery, service gateway, service arrangement, API management, an integration framework, distributed management and a call chain; the platform layer also provides traffic AI algorithm support and security authentication;
the scene layer provides calculation support of intelligent networking edge scenes including traffic safety, traffic efficiency, travel service and automatic driving, and can be elastically expanded according to requirements;
The service application layer is cooperated with the cloud service application layer on the basis of scene layer service, and provides intelligent networking and intelligent traffic service applications including bus priority, holographic intersections, traffic sequencing, digital twinning, signal optimization and free flow charging for traffic managers, traffic transportation industries and driver users;
the distributed resource allocation and scheduling policy includes:
step 1, determining business application, and defining the related traffic participation object and the action range of the business application;
dividing the edge scenes, and dividing the determined business application into different edge scene application combinations;
step 3, analyzing detailed characteristics of the edge scene application, wherein the detailed characteristics comprise application properties, event types, high-precision map and positioning requirements, time delay requirements, reliability requirements, transmission bandwidth requirements, message periods, data packet sizes, data types, AI requirements, multi-source fusion requirements, decision and control requirements, road characteristics, target motion characteristics, static characteristics, environment and weather influences of the edge scene application;
step 4, determining an edge computing task, and determining the edge computing task for realizing the edge scene according to the detailed characteristics of the edge scene application;
Step 5, determining the detailed description of the edge computing resources, submitting an edge computing task request to a platform layer, calling a corresponding micro-service and algorithm model by the platform layer through a micro-service API interface gateway according to the request, and issuing the detailed description of the required computing resources to a resource layer through the micro-service;
and 6, carrying out distributed resource scheduling, carrying out fine scheduling of computing power, bandwidth and storage according to the detailed description of the required edge computing resources by a resource layer, and distributing the proper resources to the proper edge computing tasks.
2. The urban road C-V2X network-oriented edge computing system according to claim 1, wherein the dedicated vehicle-mounted MEC terminal comprises a lightweight edge computing device dedicated to vehicle-sensed data analysis and simple fusion computation, the lightweight edge computing device being provided with an external interface through which it interfaces with a vehicle-mounted camera, millimeter wave radar, lidar, ultrasonic radar, CAN bus, beidou/GPS positioning and vehicle-mounted OBU device; the computing power support of the lightweight edge computing equipment fuses the original data perceived by the butted vehicle-mounted camera, the millimeter wave radar, the laser radar, the ultrasonic radar, the CAN bus and the Beidou/GPS positioning equipment to finish real-time traffic target detection and identification, wherein the real-time traffic target detection and identification comprises the characteristics of the distance, the speed, the direction, the acceleration, the outline, the color and the license plate number of a target, and the real-time traffic target detection and identification are transmitted to the vehicle-mounted OBU in a structured data form;
The vehicle-mounted computer comprises a vehicle-mounted special computer, wherein the vehicle-mounted special computer is installed in front of a vehicle factory, and the software and the hardware are relatively closed and are used for carrying out real-time dynamic monitoring and fault detection and early warning on the working state of the vehicle;
the vehicle-mounted OBU comprises vehicle-mounted communication equipment special for the vehicle networking, which adopts a C-V2X technology and has edge calculation capability, and the edge calculation capability of the vehicle-mounted communication equipment special for the vehicle networking can be matched with a special vehicle-mounted MEC terminal to cooperatively complete a vehicle-end edge calculation task; the vehicle-mounted OBU can interact data of the vehicle-end MEC unit with the base station MEC unit through the cellular network, and interact with other vehicle-end MEC units and road side MEC units through the V2X network.
3. The urban road C-V2X network-oriented edge computing system according to claim 1, wherein said roadside MEC unit interfaces with RSUs, roadside cameras, millimeter wave radars, lidars, intelligence boards/guidance screens, beidou/GPS, annunciators, weather sensors and environmental sensors through ethernet switches, and connects with base station MEC units through optical fibers or cellular networks.
4. An edge computing system for an urban road C-V2X network according to claim 3, wherein said edge computing tasks include latency sensitive, security sensitive, bandwidth sensitive, power consuming, memory footprint and heterogeneous collaborative tasks in task characteristics.
5. The urban road C-V2X network-oriented edge computing system according to claim 4, wherein distributed resource scheduling is performed, a scheduling scheme based on a distributed resource scheduling algorithm and a combination thereof are formulated in advance for different task characteristics of edge computing tasks by integrating the existing resource pool conditions, and a menu type resource scheduling scheme service is provided; for a vehicle-end MEC unit, computing resources of the vehicle-end MEC unit are preferentially distributed to the vehicle; for a road side MEC unit, the supported resource scheduling range is in the signal coverage range of the RSU, and is used for scene calculation including road side image analysis, laser radar point cloud processing algorithm and high-precision map matching, and supporting the provision of calculation resources from the road side to the vehicle end, so as to realize the cooperative auxiliary driving and automatic driving support of the vehicle and road; for the base station MEC unit, the scene computation is used for long period, medium delay and weak locality.
6. The urban road C-V2X network-oriented edge computing system according to claim 5, wherein said distributed resource scheduling algorithm comprises:
step 6.1, initializing a queue, dividing an edge computing task into 6 queues according to task characteristics, and recording the queue where a delay sensitive task is located as r 1 The queue where the security sensitive task is located is r 2 The queue where the bandwidth sensitive task is located is r 3 The queue where the task with power consumption is located is r 4 The queue where the memory occupied task is located is r 5 And the queue where the heterogeneous collaborative task is located is r 6 Allocating fixed length to each queue for initialization; edge computing task requests with different task characteristics enter corresponding queues in parallel；
Step 6.2, setting the priority of each queue, allocating resources to the queues with high authority preferentially, and defining the priority of each queue as r in sequence 1 >r 2 >r 3 =r 4 =r 5 =r 6 I.e. system priority handling queue r 1 Is a request in (a); in the process of receiving the request, if the queue with the highest priority reaches the allocation length, combining the queue with the next-priority queue; if a certain queue is empty in a set time interval, merging the queue with the highest priority; after the request in the highest priority queue is completely processed, processing the next priority queue;
step 6.3, determining the scheduling sequence of different requests in the same queue;
and 6.4, carrying out resource allocation processing on the requests in each queue according to the scheduling sequence.
7. The urban road C-V2X network-oriented edge computing system according to claim 6, wherein said step 6.3 comprises:
Step 6.3.1, defining QoS model indexes of the edge computing system, wherein the model indexes comprise computing time, transmission time, scheduling time, bandwidth overhead, computing power resources, storage resources, security requirements and authority indexes;
step 6.3.2 defining vectorsS={s 1 ,s 2 ,…,s i ,…,s n A set of request services in a certain queue,nto request the number of the data, 1 to less than or equal toi ≤ nThe method comprises the steps of carrying out a first treatment on the surface of the Vector quantityQ={q 1 ,q 2 ,…,q j ,…,q m Is the set of QoS model metrics,mis the total number of QoS model indexes, which is 1 to less than or equal toj ≤ mThe method comprises the steps of carrying out a first treatment on the surface of the The weight matrix isP=(p ij ) n×m ，p ij Representing the first in the queueiRequest pair numberjImportance requirements of the item QoS index;
step 6.3.3, normalizing the weight matrix to obtainTo a calculation matrixY=(y ij ) n×m ：
Wherein,,
step 6.3.4, calculate the firstiInformation entropy value of individual requestH i The method comprises the following steps:
step 6.3.5, calculate the firstiEvaluation index of individual requestsw i The method comprises the following steps:
step 6.3.6, requests to each queue are according to the evaluation indexw i And (5) arranging in a descending order to obtain the scheduling orders of different requests.
8. The edge computing system for the urban road C-V2X network according to claim 1, wherein after the distributed resource scheduling is completed in step 6, that is, after the edge computing task is completed, the corresponding container and its mirror image are deleted, and the occupied resources thereof are released in time.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111213957.0A CN113965568B (en) | 2021-10-19 | 2021-10-19 | Edge computing system for urban road C-V2X network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111213957.0A CN113965568B (en) | 2021-10-19 | 2021-10-19 | Edge computing system for urban road C-V2X network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113965568A CN113965568A (en) | 2022-01-21 |
| CN113965568B true CN113965568B (en) | 2023-07-04 |
Family
ID=79465086
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111213957.0A Active CN113965568B (en) | 2021-10-19 | 2021-10-19 | Edge computing system for urban road C-V2X network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113965568B (en) |
Families Citing this family (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114648870B (en) * | 2022-02-11 | 2023-07-28 | 行云新能科技（深圳）有限公司 | Edge computing system, edge computing decision prediction method, and computer-readable storage medium |
| CN114501378B (en) * | 2022-03-06 | 2024-08-23 | 南京理工大学 | Special environment emergency communication method and system based on vehicle-road cooperation |
| CN114637262B (en) * | 2022-03-10 | 2022-11-15 | 天津科技大学 | Decision control method and system of intelligent factory digital twin information based on 5G drive |
| CN114724391A (en) * | 2022-03-30 | 2022-07-08 | 重庆长安汽车股份有限公司 | System and method for guiding vehicles on congested road section |
| CN114625174B (en) * | 2022-05-12 | 2022-08-02 | 之江实验室 | Vehicle-mounted unmanned aerial vehicle control method and device based on V2X |
| CN115100866B (en) * | 2022-07-18 | 2023-08-18 | 北京邮电大学 | Vehicle-road cooperative automatic driving decision-making method based on layered reinforcement learning |
| CN115883660A (en) * | 2022-11-21 | 2023-03-31 | 中国联合网络通信集团有限公司 | Industrial production computing power network service method, platform, equipment and medium |
| CN116405905B (en) * | 2022-12-20 | 2024-01-30 | 联通智网科技股份有限公司 | Information processing method, device, equipment and storage medium |
| 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 |
| CN116308153B (en) * | 2023-03-01 | 2024-03-29 | 北京图安世纪科技股份有限公司 | Holographic intersection management system and method based on digital twinning |
| CN116403402B (en) * | 2023-04-13 | 2024-06-21 | 交通运输部公路科学研究所 | Traffic state prediction method for urban intersection area in network environment |
| CN116887357B (en) * | 2023-09-08 | 2023-12-19 | 山东海博科技信息系统股份有限公司 | Computing platform management system based on artificial intelligence |
| CN117939538B (en) * | 2024-03-21 | 2024-07-30 | 北京烽火万家科技有限公司 | Mobile Internet of things edge intelligent management multimode gateway |
| CN118509813B (en) * | 2024-07-22 | 2024-09-27 | 中汽智联技术有限公司 | Message processing method, equipment and system of vehicle-road cooperation automatic driving road side infrastructure |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110621006A (en) * | 2019-09-27 | 2019-12-27 | 腾讯科技（深圳）有限公司 | Access processing method of user equipment, intelligent equipment and computer storage medium |
| CN111554088A (en) * | 2020-04-13 | 2020-08-18 | 重庆邮电大学 | Multifunctional V2X intelligent roadside base station system |
| CN112289059A (en) * | 2020-10-22 | 2021-01-29 | 中电智能技术南京有限公司 | Vehicle-road cooperative road traffic system |
| CN112309122A (en) * | 2020-11-19 | 2021-02-02 | 北京清研宏达信息科技有限公司 | Intelligent bus grading decision-making system based on multi-system cooperation |
| KR20210026171A (en) * | 2019-08-29 | 2021-03-10 | 인제대학교 산학협력단 | Multi-access edge computing based Heterogeneous Networks System |
| CN113347254A (en) * | 2021-06-02 | 2021-09-03 | 安徽工程大学 | Intelligent traffic control car networking system based on V2X and control method thereof |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019156956A2 (en) * | 2018-02-06 | 2019-08-15 | Cavh Llc | Intelligent road infrastructure system (iris): systems and methods |
| US10805425B2 (en) * | 2018-10-10 | 2020-10-13 | Verizon Patent And Licensing Inc. | Method and system for edge computing network interfacing |
| US20200296054A1 (en) * | 2019-03-11 | 2020-09-17 | Wipro Limited | Method and system for providing network resource sharing in multi-access edge computing (mec) network |
-
2021
- 2021-10-19 CN CN202111213957.0A patent/CN113965568B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20210026171A (en) * | 2019-08-29 | 2021-03-10 | 인제대학교 산학협력단 | Multi-access edge computing based Heterogeneous Networks System |
| CN110621006A (en) * | 2019-09-27 | 2019-12-27 | 腾讯科技（深圳）有限公司 | Access processing method of user equipment, intelligent equipment and computer storage medium |
| CN111554088A (en) * | 2020-04-13 | 2020-08-18 | 重庆邮电大学 | Multifunctional V2X intelligent roadside base station system |
| CN112289059A (en) * | 2020-10-22 | 2021-01-29 | 中电智能技术南京有限公司 | Vehicle-road cooperative road traffic system |
| CN112309122A (en) * | 2020-11-19 | 2021-02-02 | 北京清研宏达信息科技有限公司 | Intelligent bus grading decision-making system based on multi-system cooperation |
| CN113347254A (en) * | 2021-06-02 | 2021-09-03 | 安徽工程大学 | Intelligent traffic control car networking system based on V2X and control method thereof |
Non-Patent Citations (1)
| Title |
|---|
| 网联汽车智能管控云平台设计;林凡 等;物联网技术(09);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113965568A (en) | 2022-01-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113965568B (en) | Edge computing system for urban road C-V2X network | |
| Liu et al. | High-efficiency urban traffic management in context-aware computing and 5G communication | |
| Chen et al. | Cooperative intersection management: A survey | |
| Dai et al. | A scheduling algorithm for autonomous driving tasks on mobile edge computing servers | |
| WO2016161676A1 (en) | Resource allocation system, base station, device, and method | |
| Billhardt et al. | Dynamic coordination in fleet management systems: Toward smart cyber fleets | |
| CN112925657A (en) | Vehicle road cloud cooperative processing system and method | |
| CN112466115A (en) | Bus intersection priority passing control system and method based on edge calculation | |
| WO2019156956A2 (en) | Intelligent road infrastructure system (iris): systems and methods | |
| US11585669B2 (en) | Vehicle routing using connected data analytics platform | |
| CN113835420A (en) | Function distribution system for automatic driving system | |
| CN113808389A (en) | Vehicle-road cooperation system, edge computing unit, central cloud platform and information processing method | |
| CN109993968A (en) | Traffic control system based on car networking | |
| CN112309122A (en) | Intelligent bus grading decision-making system based on multi-system cooperation | |
| Elkin et al. | IoT in traffic management: review of existing methods of road traffic regulation | |
| CN107395757B (en) | Parallel vehicle networking system based on ACP method and social physical information system | |
| CN114721806A (en) | Task scheduling and executing method and system based on digital twin | |
| Pande et al. | Dynamic service migration and resource management for vehicular clouds | |
| Jurczenia et al. | A survey of vehicular network systems for road traffic management | |
| Choudhary et al. | Novel algorithm for leader election process in virtual traffic light protocol | |
| Agarwal et al. | EMVD: efficient multitype vehicle detection algorithm using deep learning approach in vehicular communication network for radio resource management | |
| CN109857104B (en) | Unmanned technology based on road virtual rail | |
| Xia et al. | Lane scheduling around crossroads for edge computing based autonomous driving | |
| Sharma et al. | Review of recent developments in sustainable traffic management system | |
| CN111627240A (en) | Intelligent urban bus special lane facility system and control method |
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 |