WO2020168506A1 - Coordinate mapping-based multi-mode identifier network paging method and system - Google Patents

Coordinate mapping-based multi-mode identifier network paging method and system Download PDF

Info

Publication number
WO2020168506A1
WO2020168506A1 PCT/CN2019/075668 CN2019075668W WO2020168506A1 WO 2020168506 A1 WO2020168506 A1 WO 2020168506A1 CN 2019075668 W CN2019075668 W CN 2019075668W WO 2020168506 A1 WO2020168506 A1 WO 2020168506A1
Authority
WO
WIPO (PCT)
Prior art keywords
node
network
coordinate
degree
coordinates
Prior art date
Application number
PCT/CN2019/075668
Other languages
French (fr)
Chinese (zh)
Inventor
李挥
胡嘉伟
邬江兴
伊鹏
朱伏生
李文军
安辉耀
李胜飞
陈世胜
唐宏
张云勇
魏进武
陈孟尝
朱强
Original Assignee
北京大学深圳研究生院
国家数字交换***工程技术研究中心
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 北京大学深圳研究生院, 国家数字交换***工程技术研究中心 filed Critical 北京大学深圳研究生院
Priority to PCT/CN2019/075668 priority Critical patent/WO2020168506A1/en
Priority to CN201980005059.0A priority patent/CN111418192B/en
Publication of WO2020168506A1 publication Critical patent/WO2020168506A1/en

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/12—Shortest path evaluation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L61/00—Network arrangements, protocols or services for addressing or naming
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L61/00—Network arrangements, protocols or services for addressing or naming
    • H04L61/50—Address allocation

Definitions

  • the invention belongs to the field of network technology improvement, and in particular relates to a method and system for multi-mode identification network addressing based on coordinate mapping.
  • Multi-mode identification network is a new open network architecture proposed for the inherent defects of the existing Internet control capabilities being too centralized and lack of international multilateral co-management and co-governance. Specifically, it refers to the joint deployment of multiple routing identification collaborative routing addressing in networks of different architectures Network environment. For example, if a content network is deployed in a traditional network, if data can shuttle between these two types of networks, a multi-mode network environment composed of two network systems and addressed by content identification and address identification is formed. By using the advantages of different networks to work together, the multi-mode identification network can improve the basic transmission capacity of the current Internet, enhance the utilization of network resources, and enrich the network layer functions. More importantly, the multi-mode identification network reduces the dependence and limitations of the existing Internet system on address identification, and provides the possibility for multilateral co-management and co-governance of the Internet.
  • the adaptability of its routing cost to a large-scale network must be considered, and the routing cost can be measured by the scale of the forwarding table (FIB) and the number of control packets required when the network topology changes. .
  • the information center network ICN directly forwards the name of the content, while the Internet of Things (IoT) has an extremely large number of nodes, which results in extremely large addressing space and must have a high degree of In order to meet the actual needs of the network, people have to rethink how to design a routing mechanism that matches the future network.
  • Greedy Geometric Routing maps the network to a metric space and assigns an address, or coordinates, to each node in it. Each network message transmitted in the network has its destination coordinates attached to it. When forwarding, the router will calculate the geometric distance between each of its neighboring nodes and the destination, and select the one with the smallest distance as the forwarding Next hop. Since each node only needs to know the coordinate information of its neighbors, GGR can minimize the size of FIB as much as possible, thus providing a basis for us to design routing protocols for large-scale networks.
  • Hyperbolic Routing is proposed based on the scale-free nature of the network, that is, the degree of nodes in the network obeys the power distribution.
  • HR Hyperbolic Routing
  • the network is mapped to a negative curvature, that is, a hyperbolic space.
  • each node is mapped to a disk with a radius R and assigned polar coordinates ( r, ⁇ ), where the angular coordinate ⁇ represents the relative position of the node in the network, and the radius coordinate r represents the center of the node.
  • the more centralized the node the smaller the radius coordinate, that is, the closer to the center of the disk.
  • the hyperbolic distance between them will decrease as their radial coordinates decrease. Therefore, greedy routing based on hyperbolic distance will tend to select a more centralized node as its forwarding Object.
  • hyperbolic routing also has its drawbacks: Compared with traditional routing protocols based on the shortest path algorithm, the forwarding path greedily selected by hyperbolic routing has a larger transmission delay. This is not only due to the inherent disadvantages of the greedy strategy, but also It is also because most of the existing hyperbolic mapping algorithms do not consider the network delay. Since reducing delay is also one of the core goals of routing protocol design, we must tackle this problem.
  • Kleinberg is equivalent to the paper "Geographic routing using hyperbolic space” in 2007 and proposed the earliest mapping algorithm for hyperbolic routing.
  • this paper by constructing the minimum spanning tree of the network, any network can be mapped into the hyperbolic space, and based on The mapped greedy routing has a 100% success rate.
  • the algorithm must know the global topological information. At the same time, when the minimum spanning tree of the network changes, the coordinates of all nodes must be recalculated, so the adaptability to dynamic networks is poor.
  • This algorithm only considers the adjacency status of the network, and all connections are regarded as equivalent regardless of their delays. As a result, routing based on this algorithm often selects suboptimal paths with larger delays.
  • Lehman is equivalent to the 2015 “An experimental investigation of hyperbolic routing with a smart forwarding plane in ndn", trying to use NDN's adaptive and intelligent forwarding plane to delay the suboptimal path selected by the hyperbolic routing Optimization.
  • the routing node will periodically detect and record the average delay between each port and the transmission destination, and dynamically select the one with the smallest delay as the next hop.
  • the hyperbolic distance determines the interface selected at the beginning of the transmission. , And the probability of delay detection for each interface.
  • this method has a significant reduction in delay, for short-term transmission, the number of detected delay samples is not enough to optimize the forwarding path. At the same time, this method cannot optimize the worst-case scenario (that is, those paths whose delay is much greater than the theoretical optimum).
  • the purpose of the present invention is to provide a multi-mode identification network addressing method based on coordinate mapping, which aims to solve the above technical problems.
  • the present invention is realized in this way, a multi-mode identification network addressing method based on coordinate mapping.
  • the multi-mode identification network addressing method based on coordinate mapping includes the following steps:
  • d 12 cosh -1 (cosh r 1 cosh r 2 -sinh r 1 sinh r 2 cos ⁇ 12 )
  • the smallest neighbor node forwards the message
  • r and ⁇ come from the coordinates of the node, and ⁇ 12 is the central angle between the two points and the origin:
  • the further technical solution of the present invention is: multiple routing identifiers coexist in the network, such as content identifiers, identity identifiers, geographic location identifiers, and IP address identifiers, etc., through multi-identity dynamic adaptation and conversion technology, to meet various network requirements
  • the step S3 is used for addressing and routing of a variety of new types of network identifiers led by content in a large-scale network.
  • mapping in the step S1 includes angular coordinate mapping and radial coordinate mapping
  • angular coordinate mapping includes the following steps:
  • S111 Take the longitude and latitude information of the geographic location of the corner coordinates of the central node with high connectivity
  • S112. Determine whether the degree is greater than or equal to the set value for the coordinates of the non-central node with low connectivity. If so, calculate the average delay from each central node based on the IP protocol, and select the set value with the smallest delay.
  • a central node can calculate its own angular coordinates. If not, most of them have only one path to the central node, and their angular coordinates will directly copy the neighbors with the highest degree.
  • the radial coordinate mapping includes the following steps:
  • a further technical solution of the present invention is: in the step S121, the most central node in each subgraph has a radius coordinate r 0 , and only minor corrections are made to non-central nodes with a smaller radius coordinate.
  • R is the radius of the sphere
  • connection probability of two nodes is:
  • x is the hyperbolic distance between two points.
  • R can be obtained by the following integral:
  • the parameter T is the temperature, which controls the degree of aggregation of the nodes; ⁇ is the curvature of the hyperbolic space, and x′ is Hyperbolic distance between and (r,0,0);
  • Another object of the present invention is to provide a multi-mode identification network addressing system based on coordinate mapping, which includes
  • the network mapping module is used to map a scale-free multi-mode identification network to a three-dimensional hyperbolic space, and assign three-dimensional spherical coordinates to each node in the network;
  • Forwarding module used to select hyperbolic distance based on the calculated distance between nodes
  • d 12 cosh -1 (cosh r 1 cosh r 2 -sinh r 1 sinh r 2 cos ⁇ 12 )
  • the smallest neighboring node is the object of forwarding
  • r and ⁇ come from the coordinates of the node, and ⁇ 12 is the central angle between the two points and the origin:
  • the further technical solution of the present invention is: multiple routing identifiers coexist in the network, such as content identifiers, identity identifiers, geographic location identifiers, and IP address identifiers, etc., through multi-identity dynamic adaptation and conversion technology, to meet various network requirements
  • the real-time requirements of the two-zone coordinates are used for addressing and routing of a variety of new network identifications led by content in large-scale networks.
  • mapping in the network mapping module includes angular coordinate mapping and radial coordinate mapping
  • angular coordinate mapping includes
  • the central node coordinate acquisition unit is used to obtain the longitude and latitude information of the geographical position of the corner coordinates of the central node with high connectivity;
  • the non-central node coordinate acquisition unit is used to judge whether the degree of the non-central node coordinates with low connectivity is greater than or equal to the set value, if it is, then based on the IP protocol, calculate the average delay between itself and each central node, and select one of them
  • the central node with the minimum time delay is set to calculate its own angular coordinates. If not, most of them have only one path to the central node, and their angular coordinates will directly copy the neighbor with the highest degree.
  • a further technical solution of the present invention is: the radial coordinate mapping includes
  • the network division module is used to divide the global network into several sub-graphs and perform independent polar coordinate operations on each sub-graph.
  • a further technical solution of the present invention is that the most central node in each subgraph in the network division module has a radius coordinate r 0 , and only minor corrections are made to non-central nodes with a smaller radius coordinate.
  • the maximum likelihood estimation of the radius coordinate includes
  • the maximum likelihood estimation of the radius coordinate includes:
  • a priori hypothesis unit According to the statistical law of the network, we assume that the degree of the node satisfies the power distribution ⁇ ( ⁇ ) ⁇ - ⁇ , the lowest degree is ⁇ 0 , and the average degree is The relationship between the degree of the node and the diameter coordinate is: Among them, R is the radius of the sphere; according to the statistical law of the network, we assume that the connection probability of two nodes is:
  • x is the hyperbolic distance between two points.
  • R can be obtained by the following integral:
  • the parameter T is the temperature, which controls the degree of aggregation of the nodes; ⁇ is the curvature of the hyperbolic space, and x′ is Hyperbolic distance between and (r,0,0);
  • Sub-picture division and calculation unit Take m nodes i 1 , i 2 ... i m with the highest degree of centrality in the network, and the remaining nodes measure the delay between themselves and each i * . If i k is the one with the smallest delay, then the node belongs to the subgraph G k . For a node with degree ⁇ i , the maximum likelihood of its radius coordinate is estimated as:
  • the beneficial effect of the present invention is that for the traditional hyperbolic routing algorithm, the present algorithm can reduce the transmission delay by about 30% at most, while maintaining the inherent high forwarding success rate.
  • the algorithm relies on less network information, simple calculations, and easy to complete locally, thereby improving the scalability of the scheme.
  • Fig. 1 is a schematic diagram from 3 nodes to a central point provided by an embodiment of the present invention.
  • Fig. 2 is a schematic diagram of a mapping algorithm provided by an embodiment of the present invention.
  • Fig. 3 is a schematic diagram of a forwarding process provided by an embodiment of the present invention.
  • Fig. 4 is a schematic diagram of a hyperbolic disk provided by an embodiment of the present invention.
  • FIG. 5 is a schematic diagram of DS comparison between the recalculated angular coordinates and the original angular coordinates provided by an embodiment of the present invention.
  • Fig. 6 is a schematic diagram of DS comparison between recalculated diameter coordinates and original diameter coordinates provided by an embodiment of the present invention.
  • the multi-mode identification network addressing method based on coordinate mapping provided by the present invention is detailed as follows
  • delay participates in the mapping process in two ways:
  • the non-central node measures the average time delay between itself and the surrounding central node, and uses the time delay as the spherical distance to complete its own angular coordinate calculation based on the angular coordinates of the central node.
  • the global network is divided into several sub-graphs, and each sub-graph will perform its radial coordinate calculation independently to improve the locality of routing and reduce delay.
  • mapping mechanisms are used to adapt to the scale-free nature of the network.
  • nodes in the network can be roughly divided into two categories: most common user nodes with low connections, and a very small number of central nodes with very large connections . For these two types of nodes, we use different coordinate mapping mechanisms.
  • the corner coordinates of the central node come from the geographical position, and the polar coordinates come from its surroundings.
  • the non-central node uses the central node as the benchmark to measure its own coordinates.
  • This algorithm maps a scale-free network to a three-dimensional hyperbolic space
  • each node in the network is assigned three-dimensional spherical coordinates
  • Two points with The hyperbolic distance of is derived from the law of cosines:
  • ⁇ 12 is the central angle between the two points and the origin:
  • each segment of the message is attached to its destination address, and each routing node knows the coordinates of its neighbors, and greedily selects the neighbor with the smallest hyperbolic distance to the destination for forwarding .
  • Each node is first given corner coordinates To be mapped to the sphere on.
  • Spherical It can be regarded as a simulation of the earth's surface, and the corner coordinates of the node reflect its actual position in the network.
  • node i For non-central nodes with a low number of connections, we do not use the same method for them, because their location in the network depends more on local topology information than geographic information.
  • node i For node i with a degree greater than or equal to 3, based on the IP protocol (its routing is usually based on the shortest path algorithm), it will measure its average delay from each central node, and select the three central nodes j 1 with the smallest delay, j 2 ,j 3 , used to calculate own angular coordinates:
  • Equation constraint (4) represents the proportional relationship between network delay and spherical distance, and the slack variable ⁇ is used to ensure that a feasible solution can be found. As shown in Figure 1.
  • the former term of the objective function ensures that the value of the slack variable is as small as possible, and the latter term ⁇ ( ⁇ i1 + ⁇ i2 + ⁇ i3 ) is used to select the smallest sum of spherical distances when there are multiple feasible solutions.
  • the radius coordinate r reflects the center degree of the node. In a scale-free network, r should satisfy the exponential distribution.
  • the super nodes in the network may lead to the generation of delayed sub-optimal paths.
  • Shanghai has an extremely large number of Internet users. Therefore, there are several super nodes with high centrality in Shanghai.
  • the hyperbolic routing forwarding path of the message may be attracted by Shanghai’s high centrality, namely: the northern end of South Korea-Shanghai-the southern end of South Korea. Additional delay.
  • the radius coordinates are first estimated by maximum likelihood, we have the following prior conditions:
  • the degree ⁇ of the node satisfies the power distribution ⁇ ( ⁇ ) ⁇ - ⁇ , the lowest degree is ⁇ 0 , and the average degree is The degree and diameter coordinates satisfy the following relationship:
  • R is the radius of the sphere.
  • connection probability is:
  • T is the temperature, which controls the degree of aggregation of the nodes;
  • ⁇ is the curvature of the hyperbolic space; at this time, R can be obtained by the following integral:
  • ⁇ is used to adjust the relative weight of the radius coordinate and the angular coordinate in the routing process.
  • DS Delay Stretch
  • the DS comparison chart of the recalculated angle coordinates vs and the original angle coordinates respectively lists the 75th percentile (representing a bad situation) and the 95th percentile (representing almost The worst situation). It can be seen that when an appropriate central node ratio is selected, the recalculation of coordinates can better reduce the worst-case time delay. Since the 75th percentile is already close to 1, its degree of optimization is relatively small.
  • FIG. 6 it is a DS comparison chart of the recalculated diameter coordinate vs. the original diameter coordinate, where the abscissa is the number of sub-graphs divided, relative to the angular coordinates, the division of the graph and the recalculation of the diameter coordinate
  • the network delay can be reduced better.
  • the algorithm can reduce the delay by up to 30% in the worst case.
  • the hyperbolic mapping algorithm maps the network to a hyperbolic space.
  • the network is mapped to a 2-dimensional disk. It can be seen that the greater the degree of the node (which can be used to represent the popularity of the node), the closer the node is to the center of the disk, that is, the radius coordinate The smaller; and the corner coordinates of the node represent its relative position in the network. as shown in picture 2.
  • the forwarding process of hyperbolic routing The message attaches the coordinates of the destination to it, and the routing node calculates the hyperbolic distance between each next hop and the destination, and selects the smallest distance to forward. As shown in Figure 3.
  • a triangular mosaic of hyperbolic discs (Poincaré discs), where each triangle has the same size. For two points with the same visual distance, the closer to the edge of the disc, the greater the actual distance between them.
  • the three-dimensional hyperbolic sphere has similar properties. As shown in Figure 4.
  • Another object of the present invention is to provide a multi-mode identification network addressing system based on coordinate mapping, which includes
  • the network mapping module is used to map a scale-free multi-mode identification network to a three-dimensional hyperbolic space, and assign three-dimensional spherical coordinates to each node in the network;
  • Forwarding module used to select hyperbolic distance based on the calculated distance between nodes
  • d 12 cosh -1 (cosh r 1 cosh r 2 -sinh r 1 sinh r 2 cos ⁇ 12 )
  • the smallest neighboring node is the object of forwarding
  • r and ⁇ come from the coordinates of the node, and ⁇ 1 2 is the central angle between the two points and the origin:
  • routing identifications coexist in the network, such as content identification, identity identification, geospatial location identification and IP address identification, etc., through multi-identification dynamic adaptation and conversion technology, to meet the real-time needs of the network for multiple needs, including hyperbolic coordinates It is used for addressing and routing of a variety of new types of network identities led by content in large-scale networks.
  • the mapping in the network mapping module includes angular coordinate mapping and radial coordinate mapping, and the angular coordinate mapping includes
  • the central node coordinate acquisition unit is used to obtain the longitude and latitude information of the geographical position of the corner coordinates of the central node with high connectivity;
  • the non-central node coordinate acquisition unit is used to judge whether the degree of the non-central node coordinates with low connectivity is greater than or equal to the set value, if it is, then based on the IP protocol, calculate the average delay between itself and each central node, and select one of them
  • the central node with the minimum time delay is set to calculate its own angular coordinates. If not, most of them have only one path to the central node, and their angular coordinates will directly copy the neighbor with the highest degree.
  • the polar coordinate mapping includes
  • the network division module is used to divide the global network into several sub-graphs and perform independent polar coordinate operations on each sub-graph.
  • the most central node in each sub-graph in the network dividing module has a radius coordinate r 0 , and only minor corrections are made to non-central nodes with a smaller radius coordinate.
  • the maximum likelihood estimation of the radius coordinate includes
  • a priori hypothesis unit According to the statistical law of the network, we assume that the degree of the node satisfies the power distribution ⁇ ( ⁇ ) ⁇ - ⁇ , the lowest degree is ⁇ 0 , and the average degree is The relationship between the degree of the node and the diameter coordinate is: Among them, R is the radius of the sphere; according to the statistical law of the network, we assume that the connection probability of two nodes is:
  • x is the hyperbolic distance between two points.
  • R can be obtained by the following integral:
  • the parameter T is the temperature, which controls the degree of aggregation of the nodes; ⁇ is the curvature of the hyperbolic space, and x′ is Hyperbolic distance between and (r,0,0);
  • Sub-picture division and calculation unit Take m nodes i 1 , i 2 ... i m with the highest degree of centrality in the network, and the remaining nodes measure the delay between themselves and each i * . If i k is the one with the smallest delay, then the node belongs to the subgraph G k . For a node with degree ⁇ i , the maximum likelihood of its radius coordinate is estimated as:
  • this algorithm can reduce the transmission delay by up to about 30% while maintaining an inherently high forwarding success rate.
  • the algorithm relies on less network information, and the calculation is simple, easy to complete locally, thereby improving the scalability of the scheme.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The present invention is suitable for the field of network technology advancement, and provided therein is a coordinate mapping-based multi-mode identifier network paging method, comprising: S1, by means of assigning a three-dimensional world coordinate to each node in a network, mapping a multi-mode identifier network having scaleless properties to one three-dimensional hyperbolic space; S2, the sender of a message attaching node coordinates of the destination thereof to each segment of the message; and S3, when forwarding the message, a routing node calculating the hyperbolic distance between the destination and each adjacent node and selecting the smallest to serve as a forwarding object. The described algorithm relies on very little global information, and the calculation is simple and well-suited for being completed locally, thereby improving large scale network adaptability.

Description

一种基于坐标映射的多模标识网络寻址方法及***Multi-mode identification network addressing method and system based on coordinate mapping 技术领域Technical field
本发明属于网络技术改进领域，尤其涉及一种基于坐标映射的多模标识网络寻址方法及***。The invention belongs to the field of network technology improvement, and in particular relates to a method and system for multi-mode identification network addressing based on coordinate mapping.
背景技术Background technique
多模标识网络是针对现有互联网控制能力过于集中、缺乏国际性的多边共管共治等本质缺陷而提出的新型开放式网络架构，具体指不同体系结构的网络共同部署多路由标识协同路由寻址的网络环境。例如在传统网络中部署内容网络，如果数据可以在这两种网络间穿梭，便形成了一个由两种网络体系组成的、内容标识和地址标识共同寻址的多模网络环境。通过利用不同网络的优势协同工作，多模标识网络能够提升当前互联网的基础传输能力，加强网络资源的利用率，丰富网络层功能。更重要的是，多模标识网络降低了现有互联网体系对地址标识的依赖性和局限性，为互联网的多边共管共治提供可能。Multi-mode identification network is a new open network architecture proposed for the inherent defects of the existing Internet control capabilities being too centralized and lack of international multilateral co-management and co-governance. Specifically, it refers to the joint deployment of multiple routing identification collaborative routing addressing in networks of different architectures Network environment. For example, if a content network is deployed in a traditional network, if data can shuttle between these two types of networks, a multi-mode network environment composed of two network systems and addressed by content identification and address identification is formed. By using the advantages of different networks to work together, the multi-mode identification network can improve the basic transmission capacity of the current Internet, enhance the utilization of network resources, and enrich the network layer functions. More importantly, the multi-mode identification network reduces the dependence and limitations of the existing Internet system on address identification, and provides the possibility for multilateral co-management and co-governance of the Internet.
在网络协议的设计过程中，必须考虑到其路由开销对于规模化网络的适应性，而路由开销可以用转发表(FIB)的规模，和网络产生拓扑变化时所需控制报文的数量来衡量。在下一代网络中，信息中心网络(ICN)直接面向内容的名字进行转发，而物联网(IoT)则具有极其庞大的节点数目，这导致它们所需的寻址空间极其巨大，且必须具有高度的动态性以应对网络的实际需求，这不得不使人们重新思考如何设计与未来网络相匹配的路由机制。In the design process of the network protocol, the adaptability of its routing cost to a large-scale network must be considered, and the routing cost can be measured by the scale of the forwarding table (FIB) and the number of control packets required when the network topology changes. . In the next-generation network, the information center network (ICN) directly forwards the name of the content, while the Internet of Things (IoT) has an extremely large number of nodes, which results in extremely large addressing space and must have a high degree of In order to meet the actual needs of the network, people have to rethink how to design a routing mechanism that matches the future network.
贪婪几何路由(Greedy Geometric Routing,GGR)将网络映射到一个度量空间中，并为其中的每个节点赋予一个地址，亦或称作坐标。网络中传输的每段网络报文均将其目的地的坐标附于其中，路由器进行转发时，将分别计算其每个邻接节点与目的地之间的几何距离，并选择其中距离最小者作为转发的下一跳。由于每个节点只需要知悉其邻居的坐标信息，GGR可使得FIB的规模尽其可能的缩减到最小，由此为我们设计面向大规模网络的路由协议提供了基础。Greedy Geometric Routing (GGR) maps the network to a metric space and assigns an address, or coordinates, to each node in it. Each network message transmitted in the network has its destination coordinates attached to it. When forwarding, the router will calculate the geometric distance between each of its neighboring nodes and the destination, and select the one with the smallest distance as the forwarding Next hop. Since each node only needs to know the coordinate information of its neighbors, GGR can minimize the size of FIB as much as possible, thus providing a basis for us to design routing protocols for large-scale networks.
双曲路由(Hyperbolic Routing,HR)的提出基于网络的无标度性，即网络中的节点度数服从幂分布这一性质。通过映射算法，网络被映射到一个负曲率，即具有双曲性的空间中，以二维的情形为例，每个节点均被映射到具有半径R的圆盘中，并被赋予极坐标(r,θ)，其中角坐标θ代表节点在网络中的相对位置，而径坐标r则代表节点的中心程度，越是中心化的节点，其径坐标越小，即越接近圆盘的中心。当两个节点的角坐标恒定时，它们之间的双曲距离会随着它们径坐标的减少而减少，由此，基于双曲距 离的贪婪路由将倾向于选取较为中心化的节点作为其转发对象。Hyperbolic Routing (HR) is proposed based on the scale-free nature of the network, that is, the degree of nodes in the network obeys the power distribution. Through the mapping algorithm, the network is mapped to a negative curvature, that is, a hyperbolic space. Taking the two-dimensional case as an example, each node is mapped to a disk with a radius R and assigned polar coordinates ( r, θ), where the angular coordinate θ represents the relative position of the node in the network, and the radius coordinate r represents the center of the node. The more centralized the node, the smaller the radius coordinate, that is, the closer to the center of the disk. When the angular coordinates of two nodes are constant, the hyperbolic distance between them will decrease as their radial coordinates decrease. Therefore, greedy routing based on hyperbolic distance will tend to select a more centralized node as its forwarding Object.
由于现实中的大多数网络，如因特网，均具有无标度性，在采用适当映射算法的基础上，简单的基于双曲距离的贪婪策略有很大概率能将报文转发到目的节点，而对于其中转发失败的少数情形，可以采用一些辅助性的智能转发策略，使得双曲路由的成功率达到100％。Since most networks in reality, such as the Internet, are scale-free, a simple greedy strategy based on hyperbolic distance has a high probability of forwarding the message to the destination node based on the use of appropriate mapping algorithms. For the few cases where the forwarding fails, some auxiliary intelligent forwarding strategies can be adopted to make the success rate of hyperbolic routing reach 100%.
然而，双曲路由同样具有其缺陷：相较于传统上基于最短路径算法的路由协议，双曲路由贪婪选取出的转发路径具有较大的传输延迟，这不仅是出于贪婪策略的固有弊端，也是由于现有的双曲映射算法大多没有考虑网络延迟所致。由于降低延迟同样是路由协议设计的核心目标之一，我们必须着手解决这一问题。However, hyperbolic routing also has its drawbacks: Compared with traditional routing protocols based on the shortest path algorithm, the forwarding path greedily selected by hyperbolic routing has a larger transmission delay. This is not only due to the inherent disadvantages of the greedy strategy, but also It is also because most of the existing hyperbolic mapping algorithms do not consider the network delay. Since reducing delay is also one of the core goals of routing protocol design, we must tackle this problem.
Kleinberg等于2007年的"Geographic routing using hyperbolic space"一文中提出了有关双曲路由的最早映射算法，该文中，通过构造网络的最小生成树，任何网络均可以被映射到双曲空间中，且基于该映射的贪婪路由具有100％的成功率。Kleinberg is equivalent to the paper "Geographic routing using hyperbolic space" in 2007 and proposed the earliest mapping algorithm for hyperbolic routing. In this paper, by constructing the minimum spanning tree of the network, any network can be mapped into the hyperbolic space, and based on The mapped greedy routing has a 100% success rate.
该算法必须知悉全局的拓扑信息，同时，当网络的最小生成树产生了任何变化，所有节点的坐标都必须重新进行计算，因此对于动态网络的适应性较差。The algorithm must know the global topological information. At the same time, when the minimum spanning tree of the network changes, the coordinates of all nodes must be recalculated, so the adaptability to dynamic networks is poor.
Bogun,Marin,Fragkiskos Papadopoulos,Dmitri Krioukov于2010年的"Sustaining the internet with hyperbolic mapping"一文，以及Papadopoulos,Fragkiskos等于2010年的"Greedy forwarding in dynamic scalefree networks embedded in hyperbolic metric spaces"一文中，退而求其次地，不再追求将任何网络嵌入到双曲空间中，而只着眼于映射具有无标度性的那些。该方法基于有关网络的先验假设(如度数分布，连接概率等)和实际的连接状态，通过极大似然估计的方法，将网络映射到双曲空间中。在面对动态网络时，该算法所得的坐标在较长的时间内(月至年计)具有较好的稳定性，且一个新入网络的节点只需要了解其周边，而非全局的拓扑信息就可以完成其坐标计算。Bogun, Marin, Fragkiskos Papadopoulos, Dmitri Krioukov asked in the article "Sustaining the internet with hyperbolic mapping" in 2010, and Papadopoulos, Fragkiskos are equal to the 2010 "Greedy forwarding in dynamic scale free networks, retreat metric space hybrid" Secondly, we no longer pursue embedding any network into hyperbolic space, but only focus on mapping those that are scale-free. This method is based on the prior assumptions about the network (such as degree distribution, connection probability, etc.) and the actual connection state, and maps the network to the hyperbolic space through the method of maximum likelihood estimation. When facing a dynamic network, the coordinates obtained by this algorithm have good stability over a long period of time (months to years), and a new node that enters the network only needs to know its surroundings, not global topology information. It can complete its coordinate calculation.
该算法仅仅考虑网络的邻接状态，所有连接不论其延迟均被视作等价，导致基于该算法的路由经常选取出延迟较大的次优路径。This algorithm only considers the adjacency status of the network, and all connections are regarded as equivalent regardless of their delays. As a result, routing based on this algorithm often selects suboptimal paths with larger delays.
Lehman等于2015年的"An experimental investigation of hyperbolic routing with a smart forwarding plane in ndn"一文中，尝试利用NDN具有自适应性和智能性的转发平面，对双曲路由选取出的次优路径进行延迟上的优化。在传输过程中，路由节点会周期性的探测并记录每个端口与传输目的地间的平均延迟，并动态地选取延迟最小者作为下一跳，双曲距离决定了传输最开始阶段选取的接口，以及对每个接口进行延迟探测的概率。Lehman is equivalent to the 2015 "An experimental investigation of hyperbolic routing with a smart forwarding plane in ndn", trying to use NDN's adaptive and intelligent forwarding plane to delay the suboptimal path selected by the hyperbolic routing Optimization. During the transmission process, the routing node will periodically detect and record the average delay between each port and the transmission destination, and dynamically select the one with the smallest delay as the next hop. The hyperbolic distance determines the interface selected at the beginning of the transmission. , And the probability of delay detection for each interface.
尽管该方法对于延迟的减少较为显著，然而对于短期传输的情形，探测所得的延迟样本数目不足以对转发路径进行优化。同时，该方法无法优化最坏的情形(即延迟远远大于理论最优的那些路径)。Although this method has a significant reduction in delay, for short-term transmission, the number of detected delay samples is not enough to optimize the forwarding path. At the same time, this method cannot optimize the worst-case scenario (that is, those paths whose delay is much greater than the theoretical optimum).
发明内容Summary of the invention
本发明的目的在于提供一种基于坐标映射的多模标识网络寻址方法，旨在解决上述的技术问题。The purpose of the present invention is to provide a multi-mode identification network addressing method based on coordinate mapping, which aims to solve the above technical problems.
本发明是这样实现的，一种基于坐标映射的多模标识网络寻址方法，所述基于坐标映射的多模标识网络寻址方法包括以下步骤：The present invention is realized in this way, a multi-mode identification network addressing method based on coordinate mapping. The multi-mode identification network addressing method based on coordinate mapping includes the following steps:
S1、将一个具有无标度性的多模标识网络映射到一个三维双曲空间中；S1. Map a multi-mode identification network with no scale to a three-dimensional hyperbolic space;
S2、在三维双曲空间中对多模标识网络中的每个节点赋予三维球坐标；S2. Assign three-dimensional spherical coordinates to each node in the multi-mode identification network in the three-dimensional hyperbolic space;
S3、在三维双曲空间中任意节点在寻址时将其目的地的节点坐标附于每段报文中。S3. When addressing any node in the three-dimensional hyperbolic space, the node coordinates of its destination are attached to each segment of the message.
S4、根据计算的节点之间的距离选择双曲距离S4. Select the hyperbolic distance according to the calculated distance between the nodes
d 12＝cosh -1(cosh r 1cosh r 2-sinh r 1 sinh r 2cosΔθ 12) d 12 =cosh -1 (cosh r 1 cosh r 2 -sinh r 1 sinh r 2 cosΔθ 12 )
最小的相邻节点进行转发报文；The smallest neighbor node forwards the message;
其中，r、θ来自于节点的坐标，Δθ 12为两点与原点连线的中心角：
Figure PCTCN2019075668-appb-000001
Among them, r and θ come from the coordinates of the node, and Δθ 12 is the central angle between the two points and the origin:
Figure PCTCN2019075668-appb-000001
本发明的进一步技术方案是：多种路由标识并存于网络中，如内容标识、身份标识、地理空间位置标识及IP地址标识等，通过多标识动态适配及转换技术，满足网络对多种需求的实时需求，其中所述步骤S3用于在大规模网络中对以内容为首的多种新型网络标识的寻址路由。The further technical solution of the present invention is: multiple routing identifiers coexist in the network, such as content identifiers, identity identifiers, geographic location identifiers, and IP address identifiers, etc., through multi-identity dynamic adaptation and conversion technology, to meet various network requirements In the real-time demand of, the step S3 is used for addressing and routing of a variety of new types of network identifiers led by content in a large-scale network.
本发明的进一步技术方案是：所述步骤S1中的映射包括角坐标映射和径坐标映射，所述角坐标映射包括以下步骤：A further technical solution of the present invention is that the mapping in the step S1 includes angular coordinate mapping and radial coordinate mapping, and the angular coordinate mapping includes the following steps:
S111、对连接度高的中心节点角坐标取其地理位置的经纬信息；S111: Take the longitude and latitude information of the geographic location of the corner coordinates of the central node with high connectivity;
S112、对连接度较低的非中心节点坐标判断是否度数大于等于设定值，若是，则基于IP协议，测算自己距每个中心节点的平均时延，并选取其中时延最小的设定值个中心节点以计算自己的角坐标，若否，则其大多仅有一条通往中心节点的路径，其角坐标将直接复制其邻居中度数最高者。S112. Determine whether the degree is greater than or equal to the set value for the coordinates of the non-central node with low connectivity. If so, calculate the average delay from each central node based on the IP protocol, and select the set value with the smallest delay. A central node can calculate its own angular coordinates. If not, most of them have only one path to the central node, and their angular coordinates will directly copy the neighbors with the highest degree.
本发明的进一步技术方案是：所述径坐标映射包括以下步骤:A further technical solution of the present invention is: the radial coordinate mapping includes the following steps:
S121、将全局网络划分为若干子图并对每个子图进行独立的极坐标运算。S121. Divide the global network into several sub-graphs and perform independent polar coordinate operations on each sub-graph.
本发明的进一步技术方案是：所述步骤S121中每个子图中最中心的节点均具有径坐标r 0，且对于原本径坐标较小的非中心节点只进行了较小的修正。 A further technical solution of the present invention is: in the step S121, the most central node in each subgraph has a radius coordinate r 0 , and only minor corrections are made to non-central nodes with a smaller radius coordinate.
本发明的进一步技术方案是：所述径坐标的极大似然估计包括以下步骤：A further technical solution of the present invention is that the maximum likelihood estimation of the radius coordinate includes the following steps:
S1211、根据网络的统计规律，我们假定节点的度数满足幂分布ρ(κ)～κ -γ，最低度数为κ 0，平均度数为
Figure PCTCN2019075668-appb-000002
节点的度数和径坐标的关系为： S1211. According to the statistical law of the network, we assume that the degree of the node satisfies the power distribution ρ(κ)～κ -γ , the lowest degree is κ 0 , and the average degree is
Figure PCTCN2019075668-appb-000002
The relationship between the degree of the node and the diameter coordinate is:
Figure PCTCN2019075668-appb-000003
Figure PCTCN2019075668-appb-000003
其中R为球面的半径；Where R is the radius of the sphere;
S1212、根据网络的统计规律，我们假定两个节点的相连概率为：S1212. According to the statistical law of the network, we assume that the connection probability of two nodes is:
Figure PCTCN2019075668-appb-000004
Figure PCTCN2019075668-appb-000004
x是两点间的双曲距离，此时R可以由以下积分式得出：x is the hyperbolic distance between two points. At this time, R can be obtained by the following integral:
Figure PCTCN2019075668-appb-000005
Figure PCTCN2019075668-appb-000005
其中参数T为温度，控制节点的聚集程度；ζ为双曲空间的曲率，x′为
Figure PCTCN2019075668-appb-000006
和(r,0,0)之间的双曲距离； The parameter T is the temperature, which controls the degree of aggregation of the nodes; ζ is the curvature of the hyperbolic space, and x′ is
Figure PCTCN2019075668-appb-000006
Hyperbolic distance between and (r,0,0);
S1213、取网络中m个中心程度最高的节点i 1,i 2...i m，其余节点测算自己与每个i *的时延，若i k是其中时延最小者，则该节点属于子图G k。对于拥有度数k i的节点，其径坐标的极大似然估计为： S1213. Take m nodes i 1 , i 2 ... i m with the highest degree of centrality in the network, and the remaining nodes measure the delay between themselves and each i * . If i k is the one with the smallest delay, the node belongs to Subgraph G k . For a node with degree k i , the maximum likelihood of its radius coordinate is estimated as:
Figure PCTCN2019075668-appb-000007
Figure PCTCN2019075668-appb-000007
若节点i∈G j，
Figure PCTCN2019075668-appb-000008
是i j的极大似然估计值，则其径坐标为： If node i ∈ G j ,
Figure PCTCN2019075668-appb-000008
Is the maximum likelihood estimate of i j , then its radius coordinate is:
Figure PCTCN2019075668-appb-000009
Figure PCTCN2019075668-appb-000009
本发明的另一目的在于提供一种基于坐标映射的多模标识网络寻址***，所述基于坐标映射的多模标识网络寻址***包括Another object of the present invention is to provide a multi-mode identification network addressing system based on coordinate mapping, which includes
网络映射模块，用于将一个具有无标度性的多模标识网络映射到一个三维双曲空间中，对网络中的每个节点赋予三维球坐标；The network mapping module is used to map a scale-free multi-mode identification network to a three-dimensional hyperbolic space, and assign three-dimensional spherical coordinates to each node in the network;
转发模块，用于根据计算的节点之间的距离选择双曲距离Forwarding module, used to select hyperbolic distance based on the calculated distance between nodes
d 12＝cosh -1(cosh r 1cosh r 2-sinh r 1sinh r 2cosΔθ 12) d 12 =cosh -1 (cosh r 1 cosh r 2 -sinh r 1 sinh r 2 cosΔθ 12 )
最小的相邻节点作为转发的对象；The smallest neighboring node is the object of forwarding;
其中，r、θ来自于节点的坐标，Δθ 12为两点与原点连线的中心角：
Figure PCTCN2019075668-appb-000010
Among them, r and θ come from the coordinates of the node, and Δθ 12 is the central angle between the two points and the origin:
Figure PCTCN2019075668-appb-000010
本发明的进一步技术方案是：多种路由标识并存于网络中，如内容标识、身份标识、地理空间位置标识及IP地址标识等，通过多标识动态适配及转换技术，满足网络对多种需求的实时需求，其中双区坐标用于在大规模网络中对以内容为首的多种新型网络标识的寻址路由。The further technical solution of the present invention is: multiple routing identifiers coexist in the network, such as content identifiers, identity identifiers, geographic location identifiers, and IP address identifiers, etc., through multi-identity dynamic adaptation and conversion technology, to meet various network requirements The real-time requirements of the two-zone coordinates are used for addressing and routing of a variety of new network identifications led by content in large-scale networks.
本发明的进一步技术方案是：所述网络映射模块中的映射包括角坐标映射和径坐标映射，所述角坐标映射包括A further technical solution of the present invention is: the mapping in the network mapping module includes angular coordinate mapping and radial coordinate mapping, and the angular coordinate mapping includes
中心节点坐标获取单元，用于对连接度高的中心节点角坐标取其地理位置的经纬信息；The central node coordinate acquisition unit is used to obtain the longitude and latitude information of the geographical position of the corner coordinates of the central node with high connectivity;
非中心节点坐标获取单元，用于对连接度较低的非中心节点坐标判断是否度数大于等于设定值，若是，则基于IP协议，测算自己距每个中心节点的平均时延，并选取其中时延最小的设定值个中心节点以计算自己的角坐标，若否，则其大多仅有一条通往中心节点的路径，其角坐标将直接复制其邻居中度数最高者。The non-central node coordinate acquisition unit is used to judge whether the degree of the non-central node coordinates with low connectivity is greater than or equal to the set value, if it is, then based on the IP protocol, calculate the average delay between itself and each central node, and select one of them The central node with the minimum time delay is set to calculate its own angular coordinates. If not, most of them have only one path to the central node, and their angular coordinates will directly copy the neighbor with the highest degree.
本发明的进一步技术方案是：所述径坐标映射包括A further technical solution of the present invention is: the radial coordinate mapping includes
网络划分模块，用于将全局网络划分为若干子图并对每个子图进行独立的极坐标运算。The network division module is used to divide the global network into several sub-graphs and perform independent polar coordinate operations on each sub-graph.
本发明的进一步技术方案是：所述网络划分模块中每个子图中最中心的节点均具有径坐标r 0，且对于原本径坐标较小的非中心节点只进行了较小的修正。 A further technical solution of the present invention is that the most central node in each subgraph in the network division module has a radius coordinate r 0 , and only minor corrections are made to non-central nodes with a smaller radius coordinate.
本发明的进一步技术方案是：所述径坐标的极大似然估计包括A further technical solution of the present invention is: the maximum likelihood estimation of the radius coordinate includes
所述径坐标的极大似然估计包括：The maximum likelihood estimation of the radius coordinate includes:
先验假设单元。根据网络的统计规律，我们假定节点的度数满足幂分布ρ(κ)～κ -γ，最低度数为κ 0，平均度数为
Figure PCTCN2019075668-appb-000011
节点的度数和径坐标的关系为：
Figure PCTCN2019075668-appb-000012
其中R为球面的半径；根据网络的统计规律，我们假定两个节点的相连概率为： A priori hypothesis unit. According to the statistical law of the network, we assume that the degree of the node satisfies the power distribution ρ(κ)～κ -γ , the lowest degree is κ 0 , and the average degree is
Figure PCTCN2019075668-appb-000011
The relationship between the degree of the node and the diameter coordinate is:
Figure PCTCN2019075668-appb-000012
Among them, R is the radius of the sphere; according to the statistical law of the network, we assume that the connection probability of two nodes is:
Figure PCTCN2019075668-appb-000013
Figure PCTCN2019075668-appb-000013
x是两点间的双曲距离，此时R可以由以下积分式得出：x is the hyperbolic distance between two points. At this time, R can be obtained by the following integral:
Figure PCTCN2019075668-appb-000014
Figure PCTCN2019075668-appb-000014
其中参数T为温度，控制节点的聚集程度；ζ为双曲空间的曲率，x′为
Figure PCTCN2019075668-appb-000015
和(r,0,0)之间的双曲距离； The parameter T is the temperature, which controls the degree of aggregation of the nodes; ζ is the curvature of the hyperbolic space, and x′ is
Figure PCTCN2019075668-appb-000015
Hyperbolic distance between and (r,0,0);
子图划分和计算单元。取网络中m个中心程度最高的节点i 1,i 2...i m，其余节点测算自己与每个i *的时延，若i k是其中时延最小者，则该节点属于子图G k。对于拥有度数κ i的节点，其径坐标的极大似然估计为： Sub-picture division and calculation unit. Take m nodes i 1 , i 2 ... i m with the highest degree of centrality in the network, and the remaining nodes measure the delay between themselves and each i * . If i k is the one with the smallest delay, then the node belongs to the subgraph G k . For a node with degree κ i , the maximum likelihood of its radius coordinate is estimated as:
Figure PCTCN2019075668-appb-000016
Figure PCTCN2019075668-appb-000016
若节点i∈G j，
Figure PCTCN2019075668-appb-000017
是i j的极大似然估计值，则其径坐标为： If node i ∈ G j ,
Figure PCTCN2019075668-appb-000017
Is the maximum likelihood estimate of i j , then its radius coordinate is:
Figure PCTCN2019075668-appb-000018
Figure PCTCN2019075668-appb-000018
本发明的有益效果是：对于传统的双曲路由算法，本算法最多可以降低30％左右的传输延迟，同时保持了固有的高转发成功率。该算法依赖的网络信息较少，且计算简单，易于在本地完成，由此提高了该方案的可拓展性。The beneficial effect of the present invention is that for the traditional hyperbolic routing algorithm, the present algorithm can reduce the transmission delay by about 30% at most, while maintaining the inherent high forwarding success rate. The algorithm relies on less network information, simple calculations, and easy to complete locally, thereby improving the scalability of the scheme.
附图说明Description of the drawings
图1是本发明实施例提供的3节点至中心点的示意图。Fig. 1 is a schematic diagram from 3 nodes to a central point provided by an embodiment of the present invention.
图2是本发明实施例提供的映射算法示意图。Fig. 2 is a schematic diagram of a mapping algorithm provided by an embodiment of the present invention.
图3是本发明实施例提供的转发过程示意图。Fig. 3 is a schematic diagram of a forwarding process provided by an embodiment of the present invention.
图4是本发明实施例提供的双曲圆盘示意图。Fig. 4 is a schematic diagram of a hyperbolic disk provided by an embodiment of the present invention.
图5是本发明实施例提供的是重计算后的角坐标与原角坐标的DS对比示意图。FIG. 5 is a schematic diagram of DS comparison between the recalculated angular coordinates and the original angular coordinates provided by an embodiment of the present invention.
图6是本发明实施例提供的为重计算的径坐标与原径坐标的DS对比示意图。Fig. 6 is a schematic diagram of DS comparison between recalculated diameter coordinates and original diameter coordinates provided by an embodiment of the present invention.
具体实施方式detailed description
如图1-4所示，本发明提供的基于坐标映射的多模标识网络寻址方法，其详述如下As shown in Figures 1-4, the multi-mode identification network addressing method based on coordinate mapping provided by the present invention is detailed as follows
本方案的核心设计思想为如下两点：The core design ideas of this program are as follows:
1.将网络的实际延迟考虑进坐标映射算法中，以减小路由的平均时延。1. Consider the actual delay of the network into the coordinate mapping algorithm to reduce the average delay of routing.
具体而言，时延以两种方式参与进映射过程中：Specifically, delay participates in the mapping process in two ways:
1)在角坐标映射算法中，非中心节点测量自身与周边中心节点间的平均时延，并以时延作为球面距离、基于中心节点的角坐标，完成自身的角坐标计算。1) In the angular coordinate mapping algorithm, the non-central node measures the average time delay between itself and the surrounding central node, and uses the time delay as the spherical distance to complete its own angular coordinate calculation based on the angular coordinates of the central node.
2)在径坐标映射算法中，基于距离各个中心节点的时延，全局网络被分划为若干子图，每个子图会独立地进行其径坐标运算，以此提升路由的局部性，从而降低延迟。2) In the radius coordinate mapping algorithm, based on the time delay from each central node, the global network is divided into several sub-graphs, and each sub-graph will perform its radial coordinate calculation independently to improve the locality of routing and reduce delay.
2.对于网络的中心节点和非中心节点，采用不同的映射机制，以适应网络的无标度性。2. For the central and non-central nodes of the network, different mapping mechanisms are used to adapt to the scale-free nature of the network.
在无标度网络中，节点的度数分布服从幂定律，换言之，网络中的节点可以大致分为两类：大多数连接数较低的普通用户节点，和极少数具有极大连接数的中心节点。对于这两类节点，我们分别采用不同的坐标映射机制。In a scale-free network, the degree distribution of nodes obeys the power law. In other words, nodes in the network can be roughly divided into two categories: most common user nodes with low connections, and a very small number of central nodes with very large connections . For these two types of nodes, we use different coordinate mapping mechanisms.
由于中心节点的数量较少在路由过程中占有着至关重要的位置，为此 我们将其作为坐标映射运算的锚点，中心节点的角坐标来自于地理位置，而极坐标来自于其在周边区域的中心程度。而非中心节点以中心节点为基准，测算自身的坐标。Since the number of central nodes is very small, it occupies a crucial position in the routing process. For this reason, we use it as the anchor point of the coordinate mapping operation. The corner coordinates of the central node come from the geographical position, and the polar coordinates come from its surroundings. The centrality of the area. The non-central node uses the central node as the benchmark to measure its own coordinates.
本发明旨在解决的技术问题为如下两点：The technical problems that the present invention aims to solve are as follows:
1)减小双曲路由的平均传输延迟，同时不丧失其原本的高成功率和高可扩展性。1) Reduce the average transmission delay of hyperbolic routing without losing its original high success rate and high scalability.
2)减少映射算法所需的全局信息：为了提高扩展性和适应性，坐标映射计算应尽可能的只需求局部的网络知识，且不依赖于任何复杂的全局调度过程。2) Reduce the global information required by the mapping algorithm: In order to improve the scalability and adaptability, the coordinate mapping calculation should only require local network knowledge as much as possible, and does not rely on any complex global scheduling process.
本算法将一个具有无标度性的网络映射到一个三维双曲空间
Figure PCTCN2019075668-appb-000019
中，网络中的每个节点均被赋予三维球坐标
Figure PCTCN2019075668-appb-000020
两点
Figure PCTCN2019075668-appb-000021
和
Figure PCTCN2019075668-appb-000022
的双曲距离由余弦定理导出： This algorithm maps a scale-free network to a three-dimensional hyperbolic space
Figure PCTCN2019075668-appb-000019
In, each node in the network is assigned three-dimensional spherical coordinates
Figure PCTCN2019075668-appb-000020
Two points
Figure PCTCN2019075668-appb-000021
with
Figure PCTCN2019075668-appb-000022
The hyperbolic distance of is derived from the law of cosines:
d 12＝cosh -1(cosh r 1cosh r 2-sinh r 1sinh r 2cosΔθ 12)      (1) d 12 = cosh -1 (cosh r 1 cosh r 2 -sinh r 1 sinh r 2 cosΔθ 12 ) (1)
其中Δθ 12为两点与原点连线的中心角： Where Δθ 12 is the central angle between the two points and the origin:
Figure PCTCN2019075668-appb-000023
Figure PCTCN2019075668-appb-000023
当采用双曲路由时，每段报文均将其目的地的地址附于其中，每个路由节点均知晓其邻居的坐标，并贪婪地选取和目的地之间双曲距离最小的邻居进行转发。When hyperbolic routing is used, each segment of the message is attached to its destination address, and each routing node knows the coordinates of its neighbors, and greedily selects the neighbor with the smallest hyperbolic distance to the destination for forwarding .
我们的算法分为角坐标映射和径坐标映射两个部分，具体流程如下所述：Our algorithm is divided into two parts: angular coordinate mapping and radial coordinate mapping. The specific process is as follows:
1.角坐标映射1. Angular coordinate mapping
每个节点首先被赋予角坐标
Figure PCTCN2019075668-appb-000024
以此被映射到球面
Figure PCTCN2019075668-appb-000025
上。球面
Figure PCTCN2019075668-appb-000026
可以视作对地球表面的模拟，而节点的角坐标反应了其在网络中的实际位置。 Each node is first given corner coordinates
Figure PCTCN2019075668-appb-000024
To be mapped to the sphere
Figure PCTCN2019075668-appb-000025
on. Spherical
Figure PCTCN2019075668-appb-000026
It can be regarded as a simulation of the earth's surface, and the corner coordinates of the node reflect its actual position in the network.
对于那些连接度很高的节点，它们的角坐标直接来自于其地理位置，即经纬度信息，该做法出于以下理由：1)根据以往的经验，两个节点间的传输延迟正比于两者间的地理距离，基于地理位置的映射对于延迟具有较好的优化效果。2)该映射方法十分直接，便于计算。3)该映射不依赖网络的拓扑信息，因此在动态的网络环境中具有较好的稳定性。For those nodes with a high degree of connectivity, their angular coordinates are directly derived from their geographic location, that is, latitude and longitude information. This approach is based on the following reasons: 1) According to past experience, the transmission delay between two nodes is proportional to the difference between them. The geographic distance, the mapping based on geographic location has a better optimization effect for delay. 2) The mapping method is very straightforward and easy to calculate. 3) The mapping does not depend on the topology information of the network, so it has better stability in a dynamic network environment.
对于连接数较低的非中心节点，我们并不对其采用同样的方法，因为相比于地理信息，它们在网络中的位置更多的取决于局部的拓扑信息。对于度数大于等于3的节点i，基于IP协议(其路由通常基于最短路径算法)，它将测算自己距每个中心节点的平均时延，并选取其中时延最小的三个中心节点j 1,j 2,j 3，用以计算自己的角坐标： For non-central nodes with a low number of connections, we do not use the same method for them, because their location in the network depends more on local topology information than geographic information. For node i with a degree greater than or equal to 3, based on the IP protocol (its routing is usually based on the shortest path algorithm), it will measure its average delay from each central node, and select the three central nodes j 1 with the smallest delay, j 2 ,j 3 , used to calculate own angular coordinates:
若i距j k的时延为t k，j k具有角坐标
Figure PCTCN2019075668-appb-000027
i的角坐标
Figure PCTCN2019075668-appb-000028
由如下优化问题得出： If the time delay between i and j k is t k , j k has angular coordinates
Figure PCTCN2019075668-appb-000027
the angular coordinate of i
Figure PCTCN2019075668-appb-000028
Resulted from the following optimization problem:
Figure PCTCN2019075668-appb-000029
Figure PCTCN2019075668-appb-000029
s.t.λΔθ ik＝t k-ξ(k＝1,2,3)           (4) stλΔθ ik = t k -ξ(k=1, 2, 3) (4)
Δθ ik为
Figure PCTCN2019075668-appb-000030
与
Figure PCTCN2019075668-appb-000031
的中心角，由(2)得出。等式约束(4)代表网络延迟和球面距离的正比例关系，松弛变量ξ用以保证可以找到可行解。如图1所示。 Δθ ik is
Figure PCTCN2019075668-appb-000030
versus
Figure PCTCN2019075668-appb-000031
The center angle of is obtained by (2). Equation constraint (4) represents the proportional relationship between network delay and spherical distance, and the slack variable ξ is used to ensure that a feasible solution can be found. As shown in Figure 1.
目标函数的前项保证松弛变量的值尽可能小，后项ε(Δθ i1+Δθ i2+Δθ i3)则用于在存在多个可行解时，选取球面距离总和最小者。 The former term of the objective function ensures that the value of the slack variable is as small as possible, and the latter term ε (Δθ i1 + Δθ i2 + Δθ i3 ) is used to select the smallest sum of spherical distances when there are multiple feasible solutions.
对于度数小于等于2的非中心节点，由于其大多仅有一条通往中心节点的路径，其角坐标将直接复制其邻居中度数最高者。For non-central nodes with a degree less than or equal to 2, since most of them have only one path to the central node, their angular coordinates will directly copy the neighbors with the highest degree.
2.径坐标映射2. Radial coordinate mapping
径坐标r反映节点的中心程度，在无标度网络中，r应满足指数分布。The radius coordinate r reflects the center degree of the node. In a scale-free network, r should satisfy the exponential distribution.
网络中的超级节点有可能导致延迟次优路径的产生，举例而言，上海市拥有极其庞大的互联网用户数目，因此上海市存在若干中心度高的超级节点，对于一段从韩国北端发往韩国南端的报文，因为韩国的人口数目相比于上海较少，该报文的双曲路由转发路径有可能被上海的高中心度吸引，即：韩国北端——上海——韩国南端，从而导致了额外的延迟。The super nodes in the network may lead to the generation of delayed sub-optimal paths. For example, Shanghai has an extremely large number of Internet users. Therefore, there are several super nodes with high centrality in Shanghai. For a section from the northern end of South Korea to the southern end of South Korea Because the population of South Korea is smaller than that of Shanghai, the hyperbolic routing forwarding path of the message may be attracted by Shanghai’s high centrality, namely: the northern end of South Korea-Shanghai-the southern end of South Korea. Additional delay.
我们采用将全局网络分划为子图的方式以应对这一问题，每个子图中最中心的节点都具有相近的径坐标，因此，转发过程将更倾向于选择本子图中的中心节点，由此提升了路由的局部性，从而降低了传输延迟。取网络中m个中心程度最高的节点i 1,i 2...i m，其余节点测算自己与每个i *的时延，若i k是其中时延最小者，则该节点属于子图G k。 We adopt the method of dividing the global network into subgraphs to deal with this problem. The most central node in each subgraph has similar radius coordinates. Therefore, the forwarding process will be more inclined to choose the central node in this subgraph. This improves the locality of routing, thereby reducing transmission delay. Take m nodes i 1 , i 2 ... i m with the highest degree of centrality in the network, and the remaining nodes measure the delay between themselves and each i * . If i k is the one with the smallest delay, then the node belongs to the subgraph G k .
径坐标首先由极大似然估计得出，我们有如下先验条件：The radius coordinates are first estimated by maximum likelihood, we have the following prior conditions:
1)节点的度数κ满足幂分布ρ(κ)～κ -γ，最低度数为κ 0，平均度数为
Figure PCTCN2019075668-appb-000032
度数和径坐标满足如下的关系： 1) The degree κ of the node satisfies the power distribution ρ(κ)～κ -γ , the lowest degree is κ 0 , and the average degree is
Figure PCTCN2019075668-appb-000032
The degree and diameter coordinates satisfy the following relationship:
Figure PCTCN2019075668-appb-000033
Figure PCTCN2019075668-appb-000033
其中R为球面的半径。Where R is the radius of the sphere.
2)双曲距离为x的两个节点，其相连概率为：2) Two nodes with hyperbolic distance x, the connection probability is:
Figure PCTCN2019075668-appb-000034
Figure PCTCN2019075668-appb-000034
T为温度，控制节点的聚集程度；ζ为双曲空间的曲率；此时R可以由以下积分式得出：T is the temperature, which controls the degree of aggregation of the nodes; ζ is the curvature of the hyperbolic space; at this time, R can be obtained by the following integral:
Figure PCTCN2019075668-appb-000035
Figure PCTCN2019075668-appb-000035
其中x为
Figure PCTCN2019075668-appb-000036
和(r,0,0)之间的双曲距离。 Where x is
Figure PCTCN2019075668-appb-000036
The hyperbolic distance between and (r,0,0).
在给定了以上先验条件的基础上，对于拥有度数κ i的节点i，其径坐标 的极大似然估计为： Given above priori basis, has a degree κ i to node i, coordinates its diameter is maximum likelihood estimation:
Figure PCTCN2019075668-appb-000037
Figure PCTCN2019075668-appb-000037
若节点i∈G j，则其径坐标为： If the node i∈G j , then its radius coordinate is:
Figure PCTCN2019075668-appb-000038
Figure PCTCN2019075668-appb-000038
β用于调节径坐标和角坐标在路由过程中的相对权重，通过上式，每个子图中的最中心节点均具有径坐标r 0，且对于原本径坐标较小的非中心节点只进行了较小的修正。 β is used to adjust the relative weight of the radius coordinate and the angular coordinate in the routing process. Through the above formula, the most central node in each subgraph has the radius coordinate r 0 , and only the non-central nodes with smaller diameter coordinates are Minor corrections.
为评估以上算法的相对优势，我们用随机网络生成器产生了一系列满足无标度性的拓扑图以用于该算法的模拟测试。其中，节点的度数服从幂分布，而经纬度服从球面上的一致均匀分布。生成器的算法参数来自对于现实互联网的拟合，且图中两个节点的传输时延正比于它们之间的球面距离。In order to evaluate the relative advantages of the above algorithms, we used a random network generator to generate a series of topological graphs that satisfy the scale-free performance for the simulation test of the algorithm. Among them, the degree of the node obeys the power distribution, and the latitude and longitude obey the uniform distribution on the sphere. The algorithm parameters of the generator come from the fitting of the real Internet, and the transmission delay of the two nodes in the figure is proportional to the spherical distance between them.
我们随机选取网络中的两个节点作为路由的起止点，并尝试使用双曲路由将报文从起始点转发到终点。我们使用延迟伸展(DS，Delay Stretch)作为衡量算法效果的指标：一次路由测试的DS为(实际的传输延迟/理论上的最小传输延迟)。我们生成了50个随机图，对于每组实验参数，在每个图中进行了十万次随机路由测试。We randomly select two nodes in the network as the start and end points of the route, and try to use hyperbolic routing to forward the message from the start point to the end point. We use Delay Stretch (DS) as an indicator to measure the effect of the algorithm: the DS of a routing test is (the actual transmission delay/the theoretical minimum transmission delay). We generated 50 random graphs. For each set of experimental parameters, 100,000 random routing tests were performed in each graph.
在实验中，我们对比了基于原始坐标(即完全基于度数和经纬度信息的坐标)和上述算法重新计算后的坐标的路由表现。由于角坐标和径坐标的映射算法相对独立，我们将分别对其进行评估。In the experiment, we compared the routing performance based on the original coordinates (that is, the coordinates based entirely on the degree and latitude and longitude information) and the coordinates recalculated by the above algorithm. Since the mapping algorithms of angular and radial coordinates are relatively independent, we will evaluate them separately.
如图5所示，重计算后的角坐标vs与原角坐标的DS对比图，分别列出了其中的第75百分位数(代表比较坏的情况)和95百分位数(代表几乎最坏的情况)。可见，在选取适当的中心节点比例时，坐标的重计算可以较好的减小最坏情形下的时延。而75百分位数由于本身就已经接近于1，其优化程度相对不大。As shown in Figure 5, the DS comparison chart of the recalculated angle coordinates vs and the original angle coordinates respectively lists the 75th percentile (representing a bad situation) and the 95th percentile (representing almost The worst situation). It can be seen that when an appropriate central node ratio is selected, the recalculation of coordinates can better reduce the worst-case time delay. Since the 75th percentile is already close to 1, its degree of optimization is relatively small.
如图6所示，为重计算的径坐标vs与原径坐标的DS对比图，其中横坐标为分划的子图数目，相对于角坐标而言，图的分划及径坐标的重计算可以更好的缩减网络的时延，如图6所示，在适当的参数选取下，对于较坏的情形算法可以缩减至多30％的时延。As shown in Figure 6, it is a DS comparison chart of the recalculated diameter coordinate vs. the original diameter coordinate, where the abscissa is the number of sub-graphs divided, relative to the angular coordinates, the division of the graph and the recalculation of the diameter coordinate The network delay can be reduced better. As shown in Figure 6, with proper parameter selection, the algorithm can reduce the delay by up to 30% in the worst case.
双曲映射算法将网络映射到一个双曲空间中。为了视觉上的直观，网络被映射到2维的圆盘中，可见，节点的度数(可以用于代表节点的流行度)越大，则节点离圆盘中心的距离就越近，即径坐标越小；而节点的角坐标代表了它在网络中的相对位置。如图2所示。The hyperbolic mapping algorithm maps the network to a hyperbolic space. For visual intuition, the network is mapped to a 2-dimensional disk. It can be seen that the greater the degree of the node (which can be used to represent the popularity of the node), the closer the node is to the center of the disk, that is, the radius coordinate The smaller; and the corner coordinates of the node represent its relative position in the network. as shown in picture 2.
双曲路由的转发过程。报文将目的地的坐标附于其中，路由节点计算每个下一跳和目的地之间的双曲距离，并选取距离最小者进行转发。如图3所示。The forwarding process of hyperbolic routing. The message attaches the coordinates of the destination to it, and the routing node calculates the hyperbolic distance between each next hop and the destination, and selects the smallest distance to forward. As shown in Figure 3.
双曲圆盘(庞加莱圆盘)的三角形镶嵌，其中每个三角形都有相同的大小。对于视觉上距离相同的两点，越接近圆盘的边缘，它们之间的实际距离就越大。三维的双曲球体具有与之相似的性质。如图4所示。A triangular mosaic of hyperbolic discs (Poincaré discs), where each triangle has the same size. For two points with the same visual distance, the closer to the edge of the disc, the greater the actual distance between them. The three-dimensional hyperbolic sphere has similar properties. As shown in Figure 4.
本发明的另一目的在于提供一种基于坐标映射的多模标识网络寻址***，所述基于坐标映射的多模标识网络寻址***包括Another object of the present invention is to provide a multi-mode identification network addressing system based on coordinate mapping, which includes
网络映射模块，用于将一个具有无标度性的多模标识网络映射到一个三维双曲空间中，对网络中的每个节点赋予三维球坐标；The network mapping module is used to map a scale-free multi-mode identification network to a three-dimensional hyperbolic space, and assign three-dimensional spherical coordinates to each node in the network;
转发模块，用于根据计算的节点之间的距离选择双曲距离Forwarding module, used to select hyperbolic distance based on the calculated distance between nodes
d 12＝cosh -1(cosh r 1cosh r 2-sinh r 1sinh r 2cosΔθ 12) d 12 =cosh -1 (cosh r 1 cosh r 2 -sinh r 1 sinh r 2 cosΔθ 12 )
最小的相邻节点作为转发的对象；The smallest neighboring node is the object of forwarding;
其中，r、θ来自于节点的坐标，Δθ1 2为两点与原点连线的中心角：
Figure PCTCN2019075668-appb-000039
Among them, r and θ come from the coordinates of the node, and Δθ1 2 is the central angle between the two points and the origin:
Figure PCTCN2019075668-appb-000039
多种路由标识并存于网络中，如内容标识、身份标识、地理空间位置标识及IP地址标识等，通过多标识动态适配及转换技术，满足网络对多种需求的实时需求，其中双曲坐标用于在大规模网络中对以内容为首的多种新型网络标识的寻址路由。Multiple routing identifications coexist in the network, such as content identification, identity identification, geospatial location identification and IP address identification, etc., through multi-identification dynamic adaptation and conversion technology, to meet the real-time needs of the network for multiple needs, including hyperbolic coordinates It is used for addressing and routing of a variety of new types of network identities led by content in large-scale networks.
所述网络映射模块中的映射包括角坐标映射和径坐标映射，所述角坐标映射包括The mapping in the network mapping module includes angular coordinate mapping and radial coordinate mapping, and the angular coordinate mapping includes
中心节点坐标获取单元，用于对连接度高的中心节点角坐标取其地理位置的经纬信息；The central node coordinate acquisition unit is used to obtain the longitude and latitude information of the geographical position of the corner coordinates of the central node with high connectivity;
非中心节点坐标获取单元，用于对连接度较低的非中心节点坐标判断是否度数大于等于设定值，若是，则基于IP协议，测算自己距每个中心节点的平均时延，并选取其中时延最小的设定值个中心节点以计算自己的角坐标，若否，则其大多仅有一条通往中心节点的路径，其角坐标将直接复制其邻居中度数最高者。The non-central node coordinate acquisition unit is used to judge whether the degree of the non-central node coordinates with low connectivity is greater than or equal to the set value, if it is, then based on the IP protocol, calculate the average delay between itself and each central node, and select one of them The central node with the minimum time delay is set to calculate its own angular coordinates. If not, most of them have only one path to the central node, and their angular coordinates will directly copy the neighbor with the highest degree.
所述极坐标映射包括The polar coordinate mapping includes
网络划分模块，用于将全局网络划分为若干子图并对每个子图进行独立的极坐标运算。The network division module is used to divide the global network into several sub-graphs and perform independent polar coordinate operations on each sub-graph.
所述网络划分模块中每个子图中最中心的节点均具有径坐标r 0，且对于原本径坐标较小的非中心节点只进行了较小的修正。 The most central node in each sub-graph in the network dividing module has a radius coordinate r 0 , and only minor corrections are made to non-central nodes with a smaller radius coordinate.
所述径坐标的极大似然估计包括The maximum likelihood estimation of the radius coordinate includes
先验假设单元。根据网络的统计规律，我们假定节点的度数满足幂分布ρ(κ)～κ -γ，最低度数为κ 0，平均度数为
Figure PCTCN2019075668-appb-000040
节点的度数和径坐标的关系 为：
Figure PCTCN2019075668-appb-000041
其中R为球面的半径；根据网络的统计规律，我们假定两个节点的相连概率为： A priori hypothesis unit. According to the statistical law of the network, we assume that the degree of the node satisfies the power distribution ρ(κ)～κ -γ , the lowest degree is κ 0 , and the average degree is
Figure PCTCN2019075668-appb-000040
The relationship between the degree of the node and the diameter coordinate is:
Figure PCTCN2019075668-appb-000041
Among them, R is the radius of the sphere; according to the statistical law of the network, we assume that the connection probability of two nodes is:
Figure PCTCN2019075668-appb-000042
Figure PCTCN2019075668-appb-000042
x是两点间的双曲距离，此时R可以由以下积分式得出：x is the hyperbolic distance between two points. At this time, R can be obtained by the following integral:
Figure PCTCN2019075668-appb-000043
Figure PCTCN2019075668-appb-000043
其中参数T为温度，控制节点的聚集程度；ζ为双曲空间的曲率，x′为
Figure PCTCN2019075668-appb-000044
和(r,0,0)之间的双曲距离； The parameter T is the temperature, which controls the degree of aggregation of the nodes; ζ is the curvature of the hyperbolic space, and x′ is
Figure PCTCN2019075668-appb-000044
Hyperbolic distance between and (r,0,0);
子图划分和计算单元。取网络中m个中心程度最高的节点i 1,i 2...i m，其余节点测算自己与每个i *的时延，若i k是其中时延最小者，则该节点属于子图G k。对于拥有度数κ i的节点，其径坐标的极大似然估计为： Sub-picture division and calculation unit. Take m nodes i 1 , i 2 ... i m with the highest degree of centrality in the network, and the remaining nodes measure the delay between themselves and each i * . If i k is the one with the smallest delay, then the node belongs to the subgraph G k . For a node with degree κ i , the maximum likelihood of its radius coordinate is estimated as:
Figure PCTCN2019075668-appb-000045
Figure PCTCN2019075668-appb-000045
若节点i∈G j，
Figure PCTCN2019075668-appb-000046
是i j的极大似然估计值，则其径坐标为： If node i ∈ G j ,
Figure PCTCN2019075668-appb-000046
Is the maximum likelihood estimate of i j , then its radius coordinate is:
Figure PCTCN2019075668-appb-000047
Figure PCTCN2019075668-appb-000047
基于模拟实验，当采用恰当的参数时，对于传统的双曲路由算法，本算法最多可以降低30％左右的传输延迟，同时保持了固有的高转发成功率。Based on simulation experiments, when appropriate parameters are used, for the traditional hyperbolic routing algorithm, this algorithm can reduce the transmission delay by up to about 30% while maintaining an inherently high forwarding success rate.
同时，该算法依赖的网络信息较少，且计算简单，易于在本地完成，由此提高了该方案的可拓展性。At the same time, the algorithm relies on less network information, and the calculation is simple, easy to complete locally, thereby improving the scalability of the scheme.
以上所述仅为本发明的较佳实施例而已，并不用以限制本发明，凡在本发明的精神和原则之内所作的任何修改、等同替换和改进等，均应包含在本发明的保护范围之内。The above descriptions are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modification, equivalent replacement and improvement made within the spirit and principle of the present invention shall be included in the protection of the present invention. Within range.

Claims (11)

  1. 一种基于坐标映射的多模标识网络寻址方法，其特征在于，所述基于坐标映射的多模标识网络寻址方法包括以下步骤：A multi-mode identification network addressing method based on coordinate mapping is characterized in that the multi-mode identification network addressing method based on coordinate mapping includes the following steps:
    S1、将一个具有无标度性的多模标识网络映射到一个三维双曲空间中；S1. Map a multi-mode identification network with no scale to a three-dimensional hyperbolic space;
    S2、在三维双曲空间中对多模标识网络中的每个节点赋予三维球坐标；S2. Assign three-dimensional spherical coordinates to each node in the multi-mode identification network in the three-dimensional hyperbolic space;
    S3、在三维双曲空间中任意节点在寻址时将其目的地的节点坐标附于每段报文中。S3. When addressing any node in the three-dimensional hyperbolic space, the node coordinates of its destination are attached to each segment of the message.
    S4、根据计算的节点之间的距离选择双曲距离S4. Select the hyperbolic distance according to the calculated distance between the nodes
    d 12＝cosh -1(cosh r 1 cosh r 2-sinh r 1 sinh r 2 cosΔθ 12) d 12 =cosh -1 (cosh r 1 cosh r 2 -sinh r 1 sinh r 2 cosΔθ 12 )
    最小的相邻节点进行转发报文；The smallest neighbor node forwards the message;
    其中，r、θ来自于节点的坐标，Δθ 12为两点与原点连线的中心角： Among them, r and θ come from the coordinates of the node, and Δθ 12 is the central angle between the two points and the origin:
    Figure PCTCN2019075668-appb-100001
    Figure PCTCN2019075668-appb-100001
  2. 根据权利要求1所述的基于坐标映射的多模标识网络寻址方法，其特征在于，所述步骤S1中的映射包括角坐标映射和径坐标映射，所述角坐标映射包括以下步骤：The multi-mode identification network addressing method based on coordinate mapping according to claim 1, wherein the mapping in step S1 includes angular coordinate mapping and radial coordinate mapping, and the angular coordinate mapping includes the following steps:
    S111、对连接度高的中心节点角坐标取其地理位置的经纬信息；S111: Take the longitude and latitude information of the geographic location of the corner coordinates of the central node with high connectivity;
    S112、对连接度较低的非中心节点坐标判断是否度数大于等于设定值，若是，则基于IP协议，测算自己距每个中心节点的平均时延，并选取其中时延最小的设定值个中心节点以计算自己的角坐标，若否，则其大多仅有一条通往中心节点的路径，其角坐标将直接复制其邻居中度数最高者。S112. Determine whether the degree is greater than or equal to the set value for the coordinates of the non-central node with low connectivity. If so, calculate the average delay from each central node based on the IP protocol, and select the set value with the smallest delay. A central node can calculate its own angular coordinates. If not, most of them have only one path to the central node, and their angular coordinates will directly copy the neighbors with the highest degree.
  3. 根据权利要求2所述的基于坐标映射的多模标识网络寻址方法，其特征在于，所述径坐标映射包括以下步骤:The multi-mode identification network addressing method based on coordinate mapping according to claim 2, wherein the radial coordinate mapping comprises the following steps:
    S121、将全局网络划分为若干子图并对每个子图进行独立的径坐标运算。S121. Divide the global network into several sub-graphs and perform independent radial coordinate calculations on each sub-graph.
  4. 根据权利要求3所述的基于坐标映射的多模标识网络寻址方法，其特征在于，所述步骤S121中每个子图中最中心的节点均具有径坐标r 0，且对于原本径坐标较小的非中心节点只进行了较小的修正。 The multi-mode identification network addressing method based on coordinate mapping according to claim 3, wherein the most central node in each sub-graph in step S121 has a radius coordinate r 0 , and the original radius coordinate is smaller Only minor corrections have been made to non-central nodes.
  5. 根据权利要求4所述的基于坐标映射的多模标识网络寻址方法，其特征在于，所述径坐标的极大似然估计包括以下步骤：The multi-mode identification network addressing method based on coordinate mapping according to claim 4, wherein the maximum likelihood estimation of the radial coordinate comprises the following steps:
    S1211、根据网络的统计规律，我们假定节点的度数满足幂分布ρ(κ)～κ -γ，最低度数为κ 0，平均度数为
    Figure PCTCN2019075668-appb-100002
    节点的度数和径坐标的关系为：     S1211. According to the statistical law of the network, we assume that the degree of the node satisfies the power distribution ρ(κ)～κ -γ , the lowest degree is κ 0 , and the average degree is
    Figure PCTCN2019075668-appb-100002
    The relationship between the degree of the node and the diameter coordinate is:
    Figure PCTCN2019075668-appb-100003
    Figure PCTCN2019075668-appb-100003
    其中R为球面的半径；Where R is the radius of the sphere;
    S1212、根据网络的统计规律，我们假定两个节点的相连概率为：S1212. According to the statistical law of the network, we assume that the connection probability of two nodes is:
    Figure PCTCN2019075668-appb-100004
    Figure PCTCN2019075668-appb-100004
    x是两点间的双曲距离，此时R可以由以下积分式得出：x is the hyperbolic distance between two points. At this time, R can be obtained by the following integral:
    Figure PCTCN2019075668-appb-100005
    Figure PCTCN2019075668-appb-100005
    其中参数T为温度，控制节点的聚集程度；ζ为双曲空间的曲率，x′为
    Figure PCTCN2019075668-appb-100006
    和(r,0,0)之间的双曲距离；     The parameter T is the temperature, which controls the degree of aggregation of the nodes; ζ is the curvature of the hyperbolic space, and x′ is
    Figure PCTCN2019075668-appb-100006
    Hyperbolic distance between and (r,0,0);
    S1213、取网络中m个中心程度最高的节点i 1,i 2...i m，其余节点测算 自己与每个i *的时延，若i k是其中时延最小者，则该节点属于子图G k。对于拥有度数κ i的节点，其径坐标的极大似然估计为： S1213. Take m nodes i 1 , i 2 ... i m with the highest degree of centrality in the network, and the remaining nodes measure the delay between themselves and each i * . If i k is the one with the smallest delay, the node belongs to Subgraph G k . For a node with degree κ i , the maximum likelihood of its radius coordinate is estimated as:
    Figure PCTCN2019075668-appb-100007
    Figure PCTCN2019075668-appb-100007
    若节点i∈G j，
    Figure PCTCN2019075668-appb-100008
    是i j的极大似然估计值，则其径坐标为：     If node i ∈ G j ,
    Figure PCTCN2019075668-appb-100008
    Is the maximum likelihood estimate of i j , then its radius coordinate is:
    Figure PCTCN2019075668-appb-100009
    Figure PCTCN2019075668-appb-100009
  6. 一种基于坐标映射的多模标识网络寻址***，其特征在于，所述基于坐标映射的多模标识网络寻址***包括A multi-mode identification network addressing system based on coordinate mapping is characterized in that the multi-mode identification network addressing system based on coordinate mapping includes
    网络映射模块，用于将一个具有无标度性的多模标识网络映射到一个三维双曲空间中，对网络中的每个节点赋予三维球坐标；The network mapping module is used to map a scale-free multi-mode identification network to a three-dimensional hyperbolic space, and assign three-dimensional spherical coordinates to each node in the network;
    转发模块，用于根据计算的节点之间的距离选择双曲距离Forwarding module, used to select hyperbolic distance based on the calculated distance between nodes
    d 12＝cosh -1(cosh r 1 cosh r 2-sinh r 1 sinh r 2 cosΔθ 12) d 12 =cosh -1 (cosh r 1 cosh r 2 -sinh r 1 sinh r 2 cosΔθ 12 )
    最小的相邻节点作为转发的对象；The smallest neighboring node is the object of forwarding;
    其中，r、θ来自于节点的坐标，Δθ 12为两点与原点连线的中心角： Among them, r and θ come from the coordinates of the node, and Δθ 12 is the central angle between the two points and the origin:
    Figure PCTCN2019075668-appb-100010
    Figure PCTCN2019075668-appb-100010
  7. 根据权利要求6所述的包含双曲路由的多模标识网络寻址***，其特征在于，多种路由标识并存于网络中，如内容标识、身份标识、地理空 间位置标识及IP地址标识等，通过多标识动态适配及转换技术，满足网络对多种需求的实时需求，其中双曲空间坐标用于在大规模网络中对以内容为首的多种新型网络标识的寻址路由。The multi-mode identification network addressing system including hyperbolic routing according to claim 6, wherein multiple routing identifications coexist in the network, such as content identification, identity identification, geospatial location identification, and IP address identification, etc. Through multi-identity dynamic adaptation and conversion technology, it can meet the real-time needs of the network for multiple needs. The hyperbolic space coordinates are used for addressing and routing of a variety of new network identities headed by content in large-scale networks.
  8. 根据权利要求7所述的基于坐标映射的多模标识网络寻址***，其特征在于，所述网络映射模块中的映射包括角坐标映射和径坐标映射，所述角坐标映射包括The multi-mode identification network addressing system based on coordinate mapping according to claim 7, wherein the mapping in the network mapping module includes angular coordinate mapping and radial coordinate mapping, and the angular coordinate mapping includes
    中心节点坐标计算单元，用于对连接度高的中心节点角坐标取其地理位置的经纬信息；The central node coordinate calculation unit is used to take the longitude and latitude information of the geographical position of the corner coordinates of the central node with high connectivity;
    非中心节点坐标计算单元，用于对连接度较低的非中心节点坐标判断是否度数大于等于设定值，若是，则基于IP协议，测算自己距每个中心节点的平均时延，并选取其中时延最小的设定值个中心节点以计算自己的角坐标，若否，则其大多仅有一条通往中心节点的路径，其角坐标将直接复制其邻居中度数最高者。The non-central node coordinate calculation unit is used for judging whether the degree of the non-central node coordinates with low connectivity is greater than or equal to the set value, if it is, then based on the IP protocol, calculate the average delay from each central node and select one of them The central node with the minimum time delay is set to calculate its own angular coordinates. If not, most of them have only one path to the central node, and their angular coordinates will directly copy the neighbors with the highest degree.
  9. 根据权利要求8所述的基于坐标映射的多模标识网络寻址***，其特征在于，所述径坐标映射包括The multi-mode identification network addressing system based on coordinate mapping of claim 8, wherein the radial coordinate mapping comprises
    网络划分模块，用于将全局网络划分为若干子图并对每个子图进行独 立的径坐标运算。The network division module is used to divide the global network into several sub-graphs and perform independent radial coordinate operations on each sub-graph.
  10. 根据权利要求9所述的基于坐标映射的多模标识网络寻址***，其特征在于，所述网络划分模块中每个子图中最中心的节点均具有径坐标r 0，且对于原本径坐标较小的非中心节点只进行了较小的修正。 The multi-mode identification network addressing system based on coordinate mapping according to claim 9, characterized in that the most central node in each sub-graph in the network division module has a radius coordinate r 0 , and the radius coordinate is lower than the original radius coordinate. Small non-central nodes have only minor corrections.
  11. 根据权利要求10所述的基于坐标映射的多模标识网络寻址***，其特征在于，所述径坐标的极大似然估计包括：The multi-mode identification network addressing system based on coordinate mapping according to claim 10, wherein the maximum likelihood estimation of the radial coordinate comprises:
    先验假设单元。根据网络的统计规律，我们假定节点的度数满足幂分布ρ(κ)～κ -γ，最低度数为κ 0，平均度数为
    Figure PCTCN2019075668-appb-100011
    节点的度数和径坐标的关系为：
    Figure PCTCN2019075668-appb-100012
    其中R为球面的半径；根据网络的统计规律，我们假定两个节点的相连概率为：     A priori hypothesis unit. According to the statistical law of the network, we assume that the degree of the node satisfies the power distribution ρ(κ)～κ -γ , the lowest degree is κ 0 , and the average degree is
    Figure PCTCN2019075668-appb-100011
    The relationship between the degree of the node and the diameter coordinate is:
    Figure PCTCN2019075668-appb-100012
    Among them, R is the radius of the sphere; according to the statistical law of the network, we assume that the connection probability of two nodes is:
    Figure PCTCN2019075668-appb-100013
    Figure PCTCN2019075668-appb-100013
    x是两点间的双曲距离，此时R可以由以下积分式得出：x is the hyperbolic distance between two points. At this time, R can be obtained by the following integral:
    Figure PCTCN2019075668-appb-100014
    Figure PCTCN2019075668-appb-100014
    其中参数T为温度，控制节点的聚集程度；ζ为双曲空间的曲率，x′为
    Figure PCTCN2019075668-appb-100015
    和(r,0,0)之间的双曲距离；     The parameter T is the temperature, which controls the degree of aggregation of the nodes; ζ is the curvature of the hyperbolic space, and x′ is
    Figure PCTCN2019075668-appb-100015
    Hyperbolic distance between and (r,0,0);
    子图划分和计算单元。取网络中m个中心程度最高的节点i 1,i 2...i m，其余节点测算自己与每个i *的时延，若i k是其中时延最小者，则该节点属于子图G k。对于拥有度数κ i的节点，其径坐标的极大似然估计为： Sub-picture division and calculation unit. Take m nodes i 1 , i 2 ... i m with the highest degree of centrality in the network, and the remaining nodes measure the delay between themselves and each i * . If i k is the one with the smallest delay, then the node belongs to the subgraph G k . For a node with degree κ i , the maximum likelihood of its radius coordinate is estimated as:
    Figure PCTCN2019075668-appb-100016
    Figure PCTCN2019075668-appb-100016
    若节点i∈G j，
    Figure PCTCN2019075668-appb-100017
    是i j的极大似然估计值，则其径坐标为：     If node i ∈ G j ,
    Figure PCTCN2019075668-appb-100017
    Is the maximum likelihood estimate of i j , then its radius coordinate is:
    Figure PCTCN2019075668-appb-100018
    Figure PCTCN2019075668-appb-100018
PCT/CN2019/075668 2019-02-21 2019-02-21 Coordinate mapping-based multi-mode identifier network paging method and system WO2020168506A1 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2019/075668 WO2020168506A1 (en) 2019-02-21 2019-02-21 Coordinate mapping-based multi-mode identifier network paging method and system
CN201980005059.0A CN111418192B (en) 2019-02-21 2019-02-21 Multi-mode identification network addressing method and system based on coordinate mapping

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2019/075668 WO2020168506A1 (en) 2019-02-21 2019-02-21 Coordinate mapping-based multi-mode identifier network paging method and system

Publications (1)

Publication Number Publication Date
WO2020168506A1 true WO2020168506A1 (en) 2020-08-27

Family

ID=71495134

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/075668 WO2020168506A1 (en) 2019-02-21 2019-02-21 Coordinate mapping-based multi-mode identifier network paging method and system

Country Status (2)

Country Link
CN (1) CN111418192B (en)
WO (1) WO2020168506A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114070773A (en) * 2021-10-13 2022-02-18 杭州电子科技大学 Space network routing strategy based on shortest path length

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112104556B (en) * 2020-11-20 2021-03-02 广东省新一代通信与网络创新研究院 Network convergence method and system
CN112565094B (en) * 2020-12-09 2023-04-07 之江实验室 Multi-mode network topology structure
CN114143084B (en) * 2021-11-30 2024-02-23 安天科技集团股份有限公司 Malicious domain name judging method and device, electronic equipment and storage medium

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102143545A (en) * 2011-03-03 2011-08-03 上海左岸芯慧电子科技有限公司 Router and packet forwarding method thereof
CN103200096A (en) * 2013-03-13 2013-07-10 南京理工大学 Heuristic routing method avoiding key nodes in complex network
WO2018098750A1 (en) * 2016-11-30 2018-06-07 深圳天珑无线科技有限公司 Message transmission method in distributed network and node
CN108712336A (en) * 2018-05-21 2018-10-26 华东师范大学 A kind of local message dynamic routing algorithm improving scales-free network transmission capacity

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102291312A (en) * 2011-09-05 2011-12-21 电子科技大学 Routing selection method based on hierarchical IP (Internet Protocol) network coordinate system
CN103152265B (en) * 2013-03-22 2017-02-15 杭州华三通信技术有限公司 Message forwarding method and equipment in SPB (Shortest Path Bridge) network
US10812365B2 (en) * 2017-02-28 2020-10-20 Northeastern University Geohyperbolic routing and addressing schemes for networks
CN107292440B (en) * 2017-06-26 2020-12-01 河海大学 Container liner route optimization method based on BA (building block) scale-free network theory

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102143545A (en) * 2011-03-03 2011-08-03 上海左岸芯慧电子科技有限公司 Router and packet forwarding method thereof
CN103200096A (en) * 2013-03-13 2013-07-10 南京理工大学 Heuristic routing method avoiding key nodes in complex network
WO2018098750A1 (en) * 2016-11-30 2018-06-07 深圳天珑无线科技有限公司 Message transmission method in distributed network and node
CN108712336A (en) * 2018-05-21 2018-10-26 华东师范大学 A kind of local message dynamic routing algorithm improving scales-free network transmission capacity

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114070773A (en) * 2021-10-13 2022-02-18 杭州电子科技大学 Space network routing strategy based on shortest path length

Also Published As

Publication number Publication date
CN111418192B (en) 2022-03-29
CN111418192A (en) 2020-07-14

Similar Documents

Publication Publication Date Title
WO2020168506A1 (en) Coordinate mapping-based multi-mode identifier network paging method and system
Bruck et al. MAP: Medial axis based geometric routing in sensor networks
Liebeherr et al. Application-layer multicasting with delaunay triangulation overlays
US20170005877A1 (en) Data object and networking node locators
WO2005055519A1 (en) Node clustering based on user interests, application requirements and network characteristics
Tsai et al. A path-connected-cluster wireless sensor network and its formation, addressing, and routing protocols
CN109104464A (en) A kind of distributed data update method towards collaboration storage under edge calculations environment
CN101437045A (en) Method for selecting transfer node of P2P system and P2P node
WO2018098750A1 (en) Message transmission method in distributed network and node
Tan et al. Convex partition of sensor networks and its use in virtual coordinate geographic routing
Hou et al. Cluster routing-based data packet backhaul prediction method in vehicular named data networking
Lu et al. Routing protocols for underwater acoustic sensor networks: A survey from an application perspective
Yang et al. A network-embedding-based approach for scalable network navigability in content-centric social IoT
WO2018098748A1 (en) Communication method in distributed network, node, and system
Wirtz et al. Mesh-DHT: A locality-based distributed look-up structure for wireless mesh networks
Tan et al. Convex partitioning of large-scale sensor networks in complex fields: Algorithms and applications
Wedaj et al. Tbdd: Territory-bound data delivery for large scale mobile sink wireless sensor networks
WO2018098749A1 (en) Message broadcasting method in distributed network and node
Bian et al. Using hierarchical location names for scalable routing and rendezvous in wireless sensor networks
Rais et al. Optimal Routing and Clustering Technique for Wireless Sensor Networks.
CN113207138B (en) SR method, system, device and medium based on multi-quality attribute
CN102035870B (en) Polar coordinate division management-based service discovery method in wireless mobile network
Zhang et al. On the utility of concave nodes in geometric processing of large-scale sensor networks
Wang Distributed Energy Balance Clustering Algorithm in Wireless Sensor Networks
Leão et al. Latency and lifetime optimization for k-anycast routing algorithm in wireless sensor networks

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19915667

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19915667

Country of ref document: EP

Kind code of ref document: A1