CN114710199A - SDN-based dynamic deployment method and system for multiple controllers of satellite network - Google Patents

Abstract

The invention discloses a satellite network multi-controller dynamic deployment method and a system based on SDN, which comprises the steps of constructing a satellite network architecture based on SDN; acquiring global load information of the LEO satellite network, and judging the load state of the LEO satellite network based on three thresholds; if the LEO satellite network is in an overall overload state, a local overload state or an underload state, making a migration strategy based on the load state of the LEO satellite network; if the LEO satellite network is in a normal state, outputting LEO satellite network information; performing dynamic migration based on a migration strategy; updating the mapping relation between a LEO satellite switch node and a LEO satellite controller node in the LEO satellite network to obtain the current LEO satellite network; judging the load state of the current LEO satellite network based on the three thresholds, and outputting the related information of the current LEO satellite network if the load state of the current LEO satellite network is in a normal state; and if the LEO satellite network is in an abnormal state, repeating the steps until the load state of the current LEO satellite network is in a normal state.

Description

A method and system for dynamic deployment of satellite network multi-controller based on SDN

technical field

The invention relates to the technical field of satellite networks, in particular to a method and system for dynamic deployment of satellite network multi-controllers based on SDN.

Background technique

Satellite network has the characteristics of huge architecture, large number and wide distribution of satellites, flexible and changeable network topology, etc. Software Defined Network (SDN) is a flexible, programmable, and centralized control new network architecture. It consists of control plane, data plane and application plane. The basic architecture of SDN is shown in Figure 1. SDN decouples the control plane and data plane, and can allocate network resources from a global perspective and formulate effective resource allocation strategies; the control plane is used as the SDN. The core component of the architecture is usually composed of one or more controllers; the data plane is composed of simple hardware forwarding devices such as switches. The data plane devices receive decision-making instructions from the upper-layer control plane through the southbound interface, and issue them according to the control plane. When the forwarding device receives the data packet, it will first check the local flow table to find the corresponding flow table entry. If the flow table entry exists and the match is successful, it will be forwarded according to the corresponding forwarding path. , if the match fails, it is encapsulated into a Packet-in request message and sent to the controller, and then the controllers communicate with each other to find the corresponding forwarding strategy, and encapsulate it into a Packet-out message and return it to the switch. In addition, the data plane will also send the current Data such as network status and statistics are returned to the control plane through the southbound interface.

Applying SDN technology to satellite networks has the advantages of simplifying satellite node functions, enhancing satellite network control capabilities, reducing network maintenance and construction costs, and realizing flexible management of network resources. The control plane can obtain a global view of the satellite network, flexibly manage and control the satellite network, and the satellite nodes can separate the control and forwarding functions , reduce the complexity of the on-board equipment; at the same time, the satellite controller node can monitor the changes of the network nodes in real time and grasp the satellite network load situation at any time.

At present, the research on multi-controller deployment of satellite network based on SDN is mainly analyzed from the perspective of static deployment of controllers and dynamic deployment of controllers; the static deployment method of multi-controller of satellite network does not consider the dynamic changes of network traffic and topology, and always It is believed that the network traffic and network structure are consistent with the initial time, which can ensure the optimal network performance in a certain state, but cannot achieve the optimal overall performance, and cannot cope with the traffic surge caused by the sudden task of the spatial information network. increase, the controller capacity is insufficient, the controller stagnates or even causes the entire network to be paralyzed, which poses a threat to the security and reliability of the network, and affects the performance of the entire network; the dynamic deployment method of satellite network multi-controller , can actively deal with the phenomenon of uneven distribution of network traffic and uneven load of the controller when the satellite data traffic changes drastically under the needs of different network users and services.

Because the current traditional SDN-based multi-controller dynamic deployment method for satellite networks cannot adapt to the current LEO satellite network topology changes when dealing with drastic changes in satellite network data traffic, some LEO satellite controller nodes may appear abnormal, such as the overall Overload state, partial overload state and underload state; when the LEO satellite controller node is overloaded as a whole or partially overloaded, the performance of the LEO satellite controller node will be impacted, and the relevant data of the LEO satellite switch node cannot be processed normally, resulting in the entire The processing performance of the satellite network is affected; when the LEO satellite controller node is under-loaded, the processing capacity of the LEO satellite controller node cannot be reasonably used, resulting in a waste of satellite resources.

SUMMARY OF THE INVENTION

In view of the above problems, an object of the present invention is to provide a method for dynamically deploying multiple controllers in a satellite network based on SDN. The method is based on the satellite network architecture of SDN. Deployed on GEO satellites, some LEO satellites are deployed from controllers; this method readjusts the division of satellite control domains according to the traffic changes of the satellite network, adopts the "three-threshold" switch migration strategy, and dynamically changes the LEO satellite switch nodes through dynamic migration. The mapping relationship between LEO satellite controller nodes and switch nodes solves the problem of uneven distribution of current satellite network traffic, and improves satellite node resource utilization, satellite network service quality and processing performance.

The second object of the present invention is to provide an SDN-based multi-controller dynamic deployment system for satellite networks.

The first technical solution adopted by the present invention is: an SDN-based multi-controller dynamic deployment method for satellite networks, comprising the following steps:

S100: Build an SDN-based satellite network architecture;

S200: Acquire global load information of the LEO satellite network, and determine the load state of the LEO satellite network based on three thresholds; the load state of the LEO satellite network includes an overall overload state, a partial overload state, an underload state, and a normal state;

S300: If the LEO satellite network is in an overall overload state, a partial overload state or an underload state, formulate a migration strategy based on the load state of the LEO satellite network; if the LEO satellite network is in a normal state, output the LEO satellite network information;

S400: Perform dynamic migration based on the migration strategy;

S500: Update the mapping relationship between the LEO satellite switch node and the LEO satellite controller node in the LEO satellite network to obtain the current LEO satellite network;

S600: Judging the load state of the current LEO satellite network based on three thresholds, if the current load state of the LEO satellite network is in a normal state, output relevant information of the current LEO satellite network; if the current load state of the LEO satellite network is not in a normal state, then The above steps S200 to S500 are repeated until the current load state of the LEO satellite network is in a normal state.

Preferably, the step S200 includes:

Determine whether the LEO satellite network load is greater than the first threshold based on the global load information of the LEO satellite network, and if it is greater than the first threshold, then determine that the LEO satellite network is in an overall overload state;

If it is less than the first threshold, it is further judged whether the load of any LEO satellite controller node is greater than the second threshold, and if it is greater than the second threshold, it is judged that the LEO satellite network is in a local overload state;

If it is less than the second threshold, it is further judged whether the load of any LEO satellite controller node is less than the third threshold, and if it is less than the third threshold, it is judged that the LEO satellite network is in an underload state;

If it is greater than the third threshold, it is determined that the LEO satellite network is in a normal state.

Preferably, the step S300 includes: if the LEO satellite network is in an overall overload state, formulating the following migration strategy:

S311 : add a LEO satellite controller node, and determine the overloaded LEO satellite controller node as the migrated LEO satellite controller node;

S312: Select the LEO satellite switch node to be migrated from the migrated LEO satellite controller nodes based on the priority mobility rate;

S313: constructing an objective function based on the migration cost of the LEO satellite network, the post-migration LEO satellite control link delay and the load balancing of the post-migration LEO satellite network, and solving the objective function to determine the position of the newly added LEO satellite controller node; And use the newly added LEO satellite controller node as the target domain.

Preferably, the step S300 includes: if the LEO satellite network is in a local overload state, formulating the following migration strategy:

S321: Screen out any overloaded LEO satellite controller node, and determine the overloaded LEO satellite controller node as the migration LEO satellite controller node;

S322: Select the LEO satellite switch node to be migrated from the migrated LEO satellite controller nodes based on the priority mobility rate;

S323: Construct an objective function based on the migration cost of the LEO satellite network, the post-migration LEO satellite control link delay, and the load balancing of the post-migration LEO satellite network, and solve the objective function to select a suitable satellite controller node as the target domain.

Preferably, the step S300 includes: if the LEO satellite network is in an underload state, formulating the following migration strategy:

S331: Screen out any underloaded LEO satellite controller node, and determine the underloaded LEO satellite controller node as the migration LEO satellite controller node;

S332: Select the LEO satellite switch node to be migrated from the migrated LEO satellite controller nodes based on the priority mobility rate;

S333: Construct an objective function based on the migration cost of the LEO satellite network, the post-migration LEO satellite control link delay, and the load balancing of the post-migration LEO satellite network, and solve the objective function to select a suitable satellite controller node as the target domain.

Preferably, the preferential mobility is expressed by the following formula:

In the formula, Q _j is the priority mobility rate of the LEO satellite switch node s _j ; λ _j is the data flow request rate of the migrated LEO satellite switch node s _j ; d _ij is the migration LEO satellite switch node s _j and the migrated LEO satellite controller node c The shortest distance between _i ; f _i is the processing capability of the LEO satellite controller c _i .

Preferably, the objective function is represented by the following formula:

W _min =α×BL+β×newcost+γ×newT _a

In the formula, W _min is the minimum value based on the migration cost of the LEO satellite network, the LEO satellite control link delay after migration, and the load balancing parameters of the LEO satellite network after migration; BL is the load balancing parameter of the LEO satellite network after migration; newcost is LEO satellite network migration cost after normalization; newT _a is the normalized value of the LEO satellite control link delay after migration; α, β, γ are the LEO satellite network migration cost, the post-migration LEO satellite Control link delay and different weights of load balancing parameters of the LEO satellite network after migration, α+β+γ=1, 0≤α,β,γ≤1.

Preferably, the objective function is solved by using the whale optimization algorithm and the simulated annealing algorithm.

Preferably, the step S400 includes:

The migration LEO satellite controller node selects the LEO satellite switch node to be migrated, deploys migration rules to the LEO satellite switch node to be migrated, the LEO satellite switch node to be migrated sends a request to the target domain, and the target domain accepts the request of the LEO satellite switch node to be migrated , migrate the LEO satellite switch node to be migrated to the target domain;

If the LEO satellite network is in an underload state, after migrating the LEO satellite switch node to be migrated to the target domain, shut down the underloaded LEO satellite controller node.

The second technical solution adopted by the present invention is: an SDN-based satellite network multi-controller dynamic deployment system, comprising a satellite network construction module, a load state determination module, a migration strategy formulation module, a migration module, an update module and a judgment module ;

The satellite network building module is used to construct an SDN-based satellite network architecture;

The load state determination module is used to obtain the global load information of the LEO satellite network, and determine the load state of the LEO satellite network based on three thresholds; the load state of the LEO satellite network includes an overall overload state, a partial overload state, an underload state and a normal state. state;

The migration strategy formulation module is configured to formulate a migration strategy based on the load state of the LEO satellite network when the LEO satellite network is in an overall overload state, a partial overload state or an underload state, and when the LEO satellite network is in a normal state In state, output LEO satellite network information;

The migration module is configured to perform dynamic migration based on the migration strategy;

The update module is used to update the mapping relationship between the LEO satellite switch node and the LEO satellite controller node in the LEO satellite network to obtain the current LEO satellite network;

The judging module is used for judging the load state of the current LEO satellite network based on three thresholds, and if the current load state of the LEO satellite network is in a normal state, outputting the relevant information of the current LEO satellite network.

The beneficial effects of the above technical solutions:

(1) A kind of SDN-based satellite network multi-controller dynamic deployment method disclosed in the present invention is based on the SDN-based satellite network architecture, the main controller in the satellite network is deployed on the ground station, the regional controller is deployed on the GEO satellite, and part of the The LEO satellite is deployed from the controller; the present invention focuses on the research on the dynamic deployment of the controller of the LEO satellite network by means of the Iridium network topology.

(2) A kind of SDN-based satellite network multi-controller dynamic deployment method disclosed in the present invention readjusts the division of satellite control domains according to the traffic changes of the satellite network, adopts the "three-threshold" switch migration strategy, and dynamically migrates the LEO satellite switches by dynamically migrating LEO satellite switches. Node, dynamically change the mapping relationship between LEO satellite controller nodes and switch nodes, solve the problem of uneven distribution of current satellite network traffic, improve satellite node resource utilization, satellite network service quality and processing performance.

(3) A kind of SDN-based satellite network multi-controller dynamic deployment method disclosed in the present invention is aimed at the deployment problem of LEO satellite controller nodes, adopts the "three threshold" switch migration strategy, and uses the whale optimization algorithm and simulation through the switch migration mechanism. The algorithm combined with the annealing algorithm aims to optimize the network migration overhead, satellite control link delay and load balancing, and dynamically changes the mapping relationship between the controller and the switch to meet the normal satellite communication requirements.

(4) The SDN-based multi-controller dynamic deployment method for a satellite network disclosed in the present invention solves the problem that when different network users, service requirements and tasks change, and the satellite network produces sudden changes in data traffic, the controllers that may appear in the satellite network The problem of dynamic deployment of controllers in the overall overload state, partial overload state and underload state.

(5) The present invention introduces SDN into the research of satellite network, adopts a distributed deployment mode for the control plane, deploys the master controller on the ground station, deploys the regional controller on the GEO satellite, and deploys the slave controller on some LEO satellites; comprehensively consider In the satellite network, LEO satellites have the characteristics of short transmission delay, large number, and rich application scenarios, which further improves the flexibility of satellite network management and can provide users with high-efficiency and high-standard services.

Description of drawings

Figure 1 is the basic architecture diagram of SDN;

2 is a schematic flowchart of a method for dynamically deploying multiple controllers in a satellite network based on SDN according to an embodiment of the present invention;

3 is a flowchart of a method for dynamically deploying multiple controllers in an SDN-based satellite network according to an embodiment of the present invention;

4 is a schematic diagram of a detailed process of a method for dynamically deploying multiple controllers in a satellite network based on SDN according to an embodiment of the present invention;

FIG. 5 is a schematic structural diagram of an SDN-based satellite network multi-controller dynamic deployment system according to an embodiment of the present invention.

Detailed ways

The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The detailed description of the following embodiments and the accompanying drawings are used to illustrate the principle of the present invention by way of example, but not to limit the scope of the present invention, that is, the present invention is not limited to the described preferred embodiments, and the scope of the present invention is defined by the claims. limited.

In the description of the present invention, it should be noted that, unless otherwise specified, the meaning of "plurality" is two or more; the terms "first", "second", etc. are only used for the purpose of description and cannot be understood For indicating or implying relative importance; for those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

Example 1

Fig. 2, Fig. 3 and Fig. 4 are an embodiment of the present invention, for LEO satellite network topology, a kind of SDN-based satellite network multi-controller dynamic deployment method is provided, which mainly includes the following steps:

S100: Build an SDN-based satellite network architecture;

The main controller in the satellite network is deployed on the ground station, the regional controller is deployed on the GEO satellite, and some LEO satellites are deployed on the slave controller; with the help of the Iridium network topology, the dynamic deployment of the controller in the LEO satellite network is focused on research;

The present invention uses the iridium network topology to study the LEO satellite network controller deployment problem. The network topology is composed of 6 polar orbits, 11 satellites on each orbit, a total of 66 LEO satellites, the orbit plane interval is 30°, and the orbit height is 781 thousand. m, the orbit inclination angle is 86.4°; the present invention regards 66 LEO satellites as LEO satellite switch nodes, and in-band deployment is to deploy the controller on the LEO satellite switch node. At this time, the LEO satellite switch node where the controller is deployed is The propagation delay can be ignored. In addition, the control link and the data link generated by this deployment method overlap, and the information transmission process does not affect each other.

S200: Acquire global load information of the LEO satellite network, and determine the load state of the LEO satellite network based on three thresholds; the load state of the LEO satellite network includes an overall overload state, a partial overload state, an underload state and a normal state;

Perform load monitoring on the LEO satellite controller node, obtain the global load information of the LEO satellite network, and judge whether the global load of the LEO satellite network is greater than the first threshold (overall overload threshold) based on the global load information of the LEO satellite network. The LEO satellite network is in an overall overload state; if it is less than the first threshold, it is further judged whether the load of any LEO satellite controller node is greater than the second threshold (local overload threshold), and if it is greater than the second threshold, it is determined that the LEO satellite network is in partial overload If it is less than the second threshold, it is further judged whether the load of any LEO satellite controller node is less than the third threshold (underload threshold). If the threshold is set, it is determined that the LEO satellite network is in a normal state, and the LEO satellite network information is output;

Among them, the first threshold is

时，则表明LEO卫星网络全局负载出现整体过载现象；When it is greater than the first threshold, that is

When , it indicates that the global load of the LEO satellite network has an overall overload phenomenon;

时，则进一步基于第二门限进行判断。When it is less than the first threshold, that is

, the judgment is further based on the second threshold.

The second threshold is _ri = 0.8,

r_i>0.8时，则表明LEO卫星网络全局负载出现局部过载现象；When it is greater than the second threshold, that is, when

When r _i > 0.8, it indicates that the global load of the LEO satellite network is partially overloaded;

r_i<0.8时，则进一步基于第三门限进行判断；When it is less than the second threshold, that is, when

When r _i <0.8, the judgment is further based on the third threshold;

Wherein, i is a satellite controller node c _i in the LEO satellite controller node set;

C= _{ c ₁ ,..,ci ,..,cm _} is the set of LEO satellite controller nodes.

The third threshold is _ri = 0.2,

r_i<0.2时，则表明LEO卫星网络存在部分LEO卫星控制器节点的欠载现象；When it is less than the third threshold, that is, when

When r _i < 0.2, it indicates that some LEO satellite controller nodes are underloaded in the LEO satellite network;

0.2<r_i<0.8时，则LEO卫星网络处于正常状态。When it is greater than the third threshold, that is, when

When 0.2<r _i <0.8, the LEO satellite network is in a normal state.

in,

为整个LEO卫星网络控制器节点的平均资源利用率，体现当前LEO卫星控制器节点的资源利用情况；m为LEO卫星控制器节点的数量。In the formula, _{ri is the ratio of the data flow request rate of all LEO satellite switch nodes within the control domain of LEO satellite controller node c i to} the processing capability of the LEO satellite controller; λ _j is the data flow request of LEO satellite switch node s _j rate; H _ij is the connection relationship between the LEO satellite controller node c _i and the LEO satellite switch node s _j ; f _i is the processing capability of the LEO satellite controller node c _i ; j is a satellite switch node in the LEO satellite switch node set s _j ; S={s ₁ ,...,s _j ,...s _n } is the set of satellite switch nodes in the LEO network;

is the average resource utilization rate of the entire LEO satellite network controller node, reflecting the current resource utilization of the LEO satellite controller node; m is the number of LEO satellite controller nodes.

S300: If the LEO satellite network is in an overall overload state, a partial overload state or an underload state, formulate a migration strategy based on the load state of the LEO satellite network; if the LEO satellite network is in a normal state, output the LEO satellite network information;

According to the load status of the LEO satellite network, the satellite switch node migration strategy is implemented, including:

①Normal status: The normal status indicates that the current global load status of the LEO satellite network is normal, and the satellite switch node does not need to be migrated at this time;

②Overall overload status: The overall overload status means that when the entire LEO satellite network is overloaded, the deployed LEO satellite controller nodes cannot process network data normally, and the entire LEO satellite network is overloaded and may appear “paralyzed”. At this moment, it is necessary to add LEO satellite controller nodes, that is, in the overall overload state, the LEO satellite controller nodes in the LEO satellite network cannot process the data generated by the entire network, and some LEO satellite controller nodes need to be added. Some satellite switch nodes in the control domain are migrated to the newly added LEO satellite control domain;

③ Local overload state: The local overload state means that the current LEO satellite network controller node can process satellite data from the overall network point of view, but some LEO satellite controller nodes are overloaded and cannot process the data successfully. At this time, it is necessary to filter out the overload. The LEO satellite controller node of the LEO satellite controller can reduce the burden of the LEO satellite controller node by migrating the LEO satellite switch node in the control domain of the overloaded satellite controller; condition, migrate the LEO satellite switch node in the control domain of the overload controller to the control domain of other LEO satellite controller nodes;

④Underload state: Starting from the resource utilization rate of the LEO satellite network, due to the occurrence of underload phenomenon, some satellite resources are wasted. At this time, consider hibernating or shutting down some LEO satellite controller nodes. Migrate the LEO satellite switch node under the LEO satellite controller node to other control domains within the scope of processing capacity; that is, by migrating the satellite switch node in the control domain of the underloaded LEO satellite controller node to the control domain of other LEO satellite controller nodes , and then shut down the LEO satellite controller node.

If the LEO satellite network is in an overall overload state, a partial overload state or an underload state, formulating a migration strategy based on the load state of the LEO satellite network includes:

(1) If the LEO satellite network is in an overall overload state, the overall overload threshold mechanism is implemented:

S311 : adding a LEO satellite controller node, and determining that the overloaded LEO satellite controller node is a migrated LEO satellite controller node;

S312: Select the LEO satellite switch node to be migrated among the satellite switch nodes in the migrated LEO satellite controller node based on the priority mobility rate;

S313: After determining the LEO satellite switch node to be migrated, construct an objective function based on the migration cost of the LEO satellite network, the post-migration LEO satellite control link delay, and the load balancing of the post-migration LEO satellite network, and use the whale optimization algorithm and simulation for the objective function A hybrid algorithm combined with an annealing algorithm is used to solve the problem to determine the optimal position of the newly added LEO satellite controller node; the newly added LEO satellite controller node is used as the target domain.

(2) If the LEO satellite network is in a local overload state, the local overload threshold mechanism is implemented:

S321: Screen out any overloaded LEO satellite controller node, and determine the overloaded LEO satellite controller node as the migration LEO satellite controller node;

S322: Select the LEO satellite switch node to be migrated among the satellite switch nodes in the migrated LEO satellite controller node based on the priority mobility rate;

S323: After determining the LEO satellite switch node to be migrated, construct an objective function based on the migration cost of the LEO satellite network, the post-migration LEO satellite control link delay, and the load balancing of the post-migration LEO satellite network, and use the whale optimization algorithm and simulation for the objective function A hybrid algorithm combined with an annealing algorithm is used to solve the problem, and an appropriate satellite controller node is selected as the target domain.

(3) If the LEO satellite network is in an underload state, the underload threshold mechanism is implemented:

S331: Screen out any underloaded LEO satellite controller node, and lock the underloaded LEO satellite controller node as the migration LEO satellite controller node;

S332: Select the LEO satellite switch node to be migrated among the satellite switch nodes in the migrated LEO satellite controller node based on the priority mobility rate;

S333: After determining the LEO satellite switch node to be migrated, construct an objective function based on the migration cost of the LEO satellite network, the post-migration LEO satellite control link delay, and the load balancing of the post-migration LEO satellite network, and use the whale optimization algorithm and simulation for the objective function A hybrid algorithm combined with an annealing algorithm is used to solve the problem, and an appropriate satellite controller node is selected as the target domain.

Wherein, after determining to migrate the LEO satellite controller node in steps S212, S222 and S232, selecting the LEO satellite switch node to be migrated among the satellite switch nodes in the migrated LEO satellite controller node based on the priority mobility rate is specifically:

After the migration of the LEO satellite controller node is determined, it is necessary to consider how to select the migration satellite switch node; when the LEO satellite controller node is under heavy load, generally speaking, it is most suitable to migrate the satellite switch node with the largest data request rate, but blindly pursue The satellite switch node with the largest request rate is not considered to be migrated. At this time, it is necessary to judge the load of the target domain of the LEO satellite controller node and the processing capacity of the LEO satellite controller node; the satellite switch node with a large request rate transmits the satellite network. On the other hand, the larger request rate increases the load pressure on the target domain of the satellite controller, and there may be a phenomenon of insufficient processing capacity; therefore, from the disturbance to the satellite network In terms of influence, the present invention preferentially selects an edge satellite switch node with a smaller migration request rate and a farther distance from the migration controller node;

The invention integrates the load of the LEO satellite controller and the link distance, and proposes the concept of the priority mobility rate of the LEO satellite switch node. The satellite switch node with a higher priority mobility rate has a higher priority and is migrated first; The priority mobility representation Q _j of the satellite switch node s _j is as follows:

In the formula, Q _j is the priority mobility rate of the LEO satellite switch node s _j ; λ _j is the data flow request rate of the migrated LEO satellite switch node s _j ; d _ij is the migration LEO satellite switch node s _j and the migrated LEO satellite controller node c The shortest distance between _i ; f _i is the processing capability of the LEO satellite controller c _i .

When the request rate of the LEO satellite switch (ie the data flow request rate λ _j of the LEO satellite switch node s _j ) is smaller, the link distance (ie the shortest distance d _ij between the LEO satellite switch node s _j and the LEO satellite controller node c _i ) The larger the value is, the larger Q _j is, indicating that the priority of s _j is higher, and the LEO satellite switch node s _j is first migrated from the LEO satellite controller node c _i .

Wherein, after determining the LEO satellite switch node to be migrated in steps S213, S223 and S233, the target domain is further selected based on the migration cost of the LEO satellite network, the post-migration LEO satellite control link delay and the load balancing of the post-migration LEO satellite network, specifically: :

The selection of the target domain is to solve the problem of where to migrate the LEO satellite switch nodes to be migrated, and to select an appropriate LEO satellite controller node to manage and control the LEO satellite switch nodes to be migrated; the selection of the target domain is mainly related to satellite network migration costs, migration The post-satellite control link delay is related to the load balancing parameters of the post-migration satellite network;

(1) Satellite network migration cost

The migration cost of the LEO satellite switch node is mainly composed of the LEO satellite network communication cost and the LEO satellite migration rule deployment cost. When the LEO satellite switch node is migrated, a migration request will be generated, and there is LEO satellite communication cost. The premise of the migration of the LEO satellite switch node It is the LEO satellite controller node that deploys the migration rules to the flow table of the migration LEO satellite switch nodes, resulting in the LEO satellite migration rule deployment overhead; that is, the LEO satellite network migration overhead is expressed as: cost=cost _com + cost _rule , where cost is the LEO satellite network migration cost, cost _com is the LEO satellite communication cost, and cost _rule is the LEO satellite migration rule deployment cost;

The overhead of LEO satellite communication mainly includes two aspects: one is the overhead generated by the establishment of the communication relationship between the LEO satellite switch node to be migrated and the LEO satellite controller node to be migrated; The communication cost between LEO satellite communication cost _com is expressed as follows:

In the formula, cost _com is the LEO satellite communication overhead; d _ij is the shortest distance between the migration LEO satellite switch node s _j and the migration LEO satellite controller node c _i ; d _wj is the migration LEO satellite switch node s _j and the target domain c The shortest distance between _w ; λ _j is the data flow request rate of migrating LEO satellite switch node s _j ;

The LEO satellite migration rule deployment cost cost _rule is expressed as:

cost _rule =ξ×d _ij ×H _ij

In the formula, cost _rule is the deployment cost of LEO satellite migration rules; ξ is the average size of flow_mod flow rule data packets; d _ij is the shortest distance between the migration LEO satellite switch node s _j and the migration LEO satellite controller node c _i ; H _ij is the connection relationship between the migrated LEO satellite controller node c _i and the migrated LEO satellite switch node s _j ; the deployment cost of the LEO satellite migration rule is only related to the migrated LEO satellite switch node.

(2) Delay of LEO satellite control link after migration

The post-migration LEO satellite control link delay mainly includes the inter-satellite link propagation delay from the LEO satellite switch node to the LEO satellite controller node in the control domain of the migrated LEO satellite controller, the network queuing delay and the LEO satellite task processing time. extend;

为LEO卫星控制器节点c_i的处理时延；

为LEO卫星控制器节点c_i与LEO卫星交换机节点间的传播时延。In the formula, T _a is the LEO satellite control link delay after migration; T _i is the total delay in the control domain of the LEO satellite controller node c _i ; Tcc is the total propagation delay between controller nodes in the entire LEO satellite network ; T _qi is the queuing delay in the control domain of the LEO satellite controller node c _i ;

is the processing delay of the LEO satellite controller node c _i ;

is the propagation delay between the LEO satellite controller node c _i and the LEO satellite switch node.

(3) Load balancing parameters of the LEO satellite network after migration

表示为：The load balancing parameters of the LEO satellite network can be used to evaluate the degree of difference in the load of the LEO satellite controller nodes. The variance method is used to calculate the difference between the load of each LEO satellite controller node and the average load of the LEO satellite controller nodes of the entire network. The closer the result is to 0, indicating that the load of each LEO satellite controller node is more balanced, and the average load of the LEO satellite controller node

Expressed as:

为LEO卫星控制器节点的平均负载；m为LEO卫星控制器节点的数量；

为LEO卫星控制器节点c_i的负载；In the formula,

is the average load of LEO satellite controller nodes; m is the number of LEO satellite controller nodes;

is the load of the LEO satellite controller node c _i ;

Then the load balancing parameters of the LEO satellite network after migration are expressed as:

为LEO卫星控制器节点的平均负载；m为LEO卫星控制器节点的数量；

为LEO卫星控制器节点c_i的负载；In the formula, BL is the load balancing parameter of the LEO satellite network after migration;

is the average load of LEO satellite controller nodes; m is the number of LEO satellite controller nodes;

is the load of the LEO satellite controller node c _i ;

The problem of selecting the target domain for migrating LEO satellite switches is the key to the migration of LEO satellite switches. It is necessary to comprehensively analyze the migration cost of the LEO satellite network, the delay of the LEO satellite control link after the migration, and the load balancing parameters of the LEO satellite network after the migration. Analysis: In order to better balance the three goals of LEO satellite network migration cost, post-migration LEO satellite control link delay, and post-migration LEO satellite network load balancing parameters, it is necessary to analyze the migration cost of LEO satellite network and the post-migration LEO satellite network migration cost. The control link delay T _a is normalized, that is:

Then the objective function W _min is defined as

W _min =α×BL+β×newcost+γ×newT _a

In the formula, W _min is the minimum value of the LEO satellite network migration overhead, the post-migration LEO satellite control link delay, and the load balancing parameters of the post-migration LEO satellite network to determine the target domain selection; BL is the post-migration LEO satellite network The load balancing parameter of the LEO satellite network; newcost is the normalized value of the LEO satellite network migration cost; newT _a is the normalized value of the LEO satellite control link delay after the migration; α, β, γ are three The numbers from 0 to 1 represent the different weights of LEO satellite network migration overhead, post-migration LEO satellite control link delay, and post-migration LEO satellite network load balancing parameters, α+β+γ=1, 0≤α,β , γ≤1.

Whale Optimization Algorithm (WOA) is a new swarm intelligence optimization algorithm. The algorithm has the advantages of simple structure, few parameters, strong search ability and easy implementation, but it is easy to fall into local optimum; considering The global search ability of the simulated annealing algorithm is strong, so the whale optimization algorithm and the simulated annealing algorithm are combined to improve the global search ability of the algorithm; therefore, after the objective function is determined, a hybrid algorithm combining the whale optimization algorithm and the simulated annealing algorithm is used. Solve the problem and select the appropriate satellite controller node as the target domain.

Further, in one embodiment, the selection of the target domain also considers the reliability of the inter-satellite link.

The invention adopts the migration strategy of "three thresholds" LEO satellite switch nodes, mainly through dynamic migration of LEO satellite switch nodes, to solve the problem of uneven distribution of current LEO satellite network traffic, including the situation of the overall overload of the LEO satellite network and the partial overload of the LEO satellite network. The situation and the situation that some satellite controller nodes are underloaded in the LEO satellite network; the dynamic deployment strategy of the LEO satellite controller node of the present invention is mainly realized by migrating the LEO satellite switch node under the LEO satellite controller node, and the LEO satellite switch Migration overhead will be generated in the process of node migration, and the load and network delay of LEO satellite controller nodes will also change before and after migration; therefore, the present invention aims at migration overhead, load balance and delay to formulate a migration strategy for LEO satellite switch nodes .

S400: Perform dynamic migration based on the migration strategy, that is, migrate the LEO satellite switch node to be migrated to the target domain;

(1) If the LEO satellite network is in an overall overload state, perform the following migration operations:

The migrated LEO satellite controller node (that is, the overloaded LEO satellite controller node) deploys migration rules to the LEO satellite switch node to be migrated, and the to-be-migrated LEO satellite switch node sends a request to the newly added LEO satellite controller node (target domain), The newly added LEO satellite controller node accepts the request of the LEO satellite switch node to be migrated, and migrates the selected LEO satellite switch node to be migrated to the newly added LEO satellite controller node to realize dynamic migration of the LEO satellite switch node.

(2) If the LEO satellite network is in a local overload state, perform the following migration operations:

Migrating the LEO satellite controller node (that is, the overloaded LEO satellite controller node) selects the LEO satellite switch node to be migrated, and deploys migration rules to the LEO satellite switch node to be migrated. The satellite controller node) sends a request, the target domain accepts the request of the LEO satellite switch node to be migrated, and migrates the selected LEO satellite switch node to be migrated to the target domain to realize the dynamic migration of the LEO satellite switch node.

(3) If the LEO satellite network is in an underload state, perform the following migration operations:

Migrating the LEO satellite controller node (ie, the underloaded LEO satellite controller node) selects the LEO satellite switch node to be migrated, and deploys migration rules to the LEO satellite switch node to be migrated. The LEO satellite controller node) sends a request, the target domain accepts the request of the LEO satellite switch node to be migrated, migrates the selected LEO satellite switch node to be migrated to the target domain, and closes the above-mentioned underloaded LEO satellite controller node. Realize dynamic migration of LEO satellite switch nodes.

S500: Update the mapping relationship between the LEO satellite switch node and the LEO satellite controller node in the LEO satellite network to obtain the current LEO satellite network;

After the LEO satellite switch node is migrated, the number and network load of the LEO satellite switch nodes in the control domain of the LEO satellite controller node will change. The mapping relationship is updated.

S600: After the LEO satellite network is updated, judge the LEO satellite controller nodes in the current LEO satellite network based on the three threshold load monitoring modules. If it is judged that the load state of the current LEO satellite network is in a normal state, output the current LEO satellite network Relevant information; if it is determined that the current load state of the LEO satellite network is not in a normal state, repeat the above steps S200 to S500 until it is determined that the current load state of the LEO satellite network is in a normal state.

Embodiment 2

5 is an SDN-based satellite network multi-controller dynamic deployment system provided by an embodiment of the present invention, including a satellite network construction module, a load state determination module, a migration strategy formulation module, a migration module, an update module, and a judgment module;

The satellite network building block is used to construct the SDN-based satellite network architecture;

The load status determination module is used to obtain the global load information of the LEO satellite network, and determine the load status of the LEO satellite network based on three thresholds; the load status of the LEO satellite network includes the overall overload status, partial overload status, underload status and normal status;

The migration strategy formulation module is configured to formulate a migration strategy based on the load state of the LEO satellite network when the LEO satellite network is in an overall overload state, a partial overload state or an underload state, and when the LEO satellite network is in a normal state , output LEO satellite network information;

The migration module is configured to perform dynamic migration based on the migration strategy;

The update module is used to update the mapping relationship between the LEO satellite switch node and the LEO satellite controller node in the LEO satellite network to obtain the current LEO satellite network;

The judging module is used for judging the load state of the current LEO satellite network based on the three thresholds, and if the current load state of the LEO satellite network is in a normal state, it outputs relevant information of the current LEO satellite network.

The following descriptions need to be made in conjunction with the present invention: First, no matter what kind of network state (overall overload state, partial overload state or underload state), the same LEO satellite switch node migration method is adopted; There is no failure phenomenon in the LEO satellite switch nodes; thirdly, the visible LEO satellites can be connected.

In the embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the apparatus embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored, or not implemented. On the other hand, the shown or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection of devices or units, and may be in electrical, mechanical or other forms.

The units described as separate components may or may not be physically separated, and components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution in this embodiment.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit.

The functions, if implemented in the form of software functional units and sold or used as independent products, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product in essence, or the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: a U disk, a removable hard disk, a ROM, a RAM, a magnetic disk, or an optical disk and other mediums that can store program codes.

The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention. should be included within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims (10)

1. a kind of satellite network multi-controller dynamic deployment method based on SDN, is characterized in that, comprises the following steps: S100: Build an SDN-based satellite network architecture; S200: Obtain global load information of the LEO satellite network, and determine the load state of the LEO satellite network based on three thresholds; the load state of the LEO satellite network includes an overall overload state, a partial overload state, an underload state, and a normal state; S300: If the LEO satellite network is in an overall overload state, a partial overload state or an underload state, formulate a migration strategy based on the load state of the LEO satellite network; if the LEO satellite network is in a normal state, output the LEO satellite network information; S400: Perform dynamic migration based on the migration strategy; S500: Update the mapping relationship between the LEO satellite switch node and the LEO satellite controller node in the LEO satellite network to obtain the current LEO satellite network; S600: Judging the load state of the current LEO satellite network based on three thresholds, if the load state of the current LEO satellite network is in a normal state, output relevant information of the current LEO satellite network; if the load state of the current LEO satellite network is not in a normal state, then The above steps S200 to S500 are repeated until the current load state of the LEO satellite network is in a normal state. 2. The method for dynamically deploying multiple controllers in a satellite network according to claim 1, wherein the step S200 comprises: Based on the global load information of the LEO satellite network, determine whether the load of the LEO satellite network is greater than the first threshold, and if it is greater than the first threshold, it is determined that the LEO satellite network is in an overall overload state; If it is less than the first threshold, it is further judged whether the load of any LEO satellite controller node is greater than the second threshold, and if it is greater than the second threshold, it is judged that the LEO satellite network is in a local overload state; If it is less than the second threshold, it is further judged whether the load of any LEO satellite controller node is less than the third threshold, and if it is less than the third threshold, it is judged that the LEO satellite network is in an underload state; If it is greater than the third threshold, it is determined that the LEO satellite network is in a normal state. 3. The satellite network multi-controller dynamic deployment method according to claim 1, wherein the step S300 comprises: if the LEO satellite network is in an overall overload state, then formulating the following migration strategy: S311 : add a LEO satellite controller node, and determine the overloaded LEO satellite controller node as the migrated LEO satellite controller node; S312: Select the LEO satellite switch node to be migrated from the migrated LEO satellite controller nodes based on the priority mobility rate; S313: constructing an objective function based on the migration cost of the LEO satellite network, the post-migration LEO satellite control link delay, and the load balancing of the post-migration LEO satellite network, and solving the objective function to determine the position of the newly added LEO satellite controller node; And use the newly added LEO satellite controller node as the target domain. 4. The satellite network multi-controller dynamic deployment method according to claim 1, wherein the step S300 comprises: if the LEO satellite network is in a local overload state, then formulating the following migration strategy: S321: Screen out any overloaded LEO satellite controller node, and determine the overloaded LEO satellite controller node as the migration LEO satellite controller node; S322: Select the LEO satellite switch node to be migrated from the migrated LEO satellite controller nodes based on the priority mobility rate; S323: Construct an objective function based on the migration cost of the LEO satellite network, the post-migration LEO satellite control link delay, and the load balancing of the post-migration LEO satellite network, and solve the objective function to select an appropriate satellite controller node as the target domain. 5. The satellite network multi-controller dynamic deployment method according to claim 1, wherein the step S300 comprises: if the LEO satellite network is in an underload state, then formulating the following migration strategy: S331: Screen out any underloaded LEO satellite controller node, and determine the underloaded LEO satellite controller node as the migration LEO satellite controller node; S332: Select the LEO satellite switch node to be migrated from the migrated LEO satellite controller nodes based on the priority mobility rate; S333: Construct an objective function based on the migration cost of the LEO satellite network, the post-migration LEO satellite control link delay, and the load balancing of the post-migration LEO satellite network, and solve the objective function to select a suitable satellite controller node as the target domain. 6. The method for dynamically deploying multiple controllers in a satellite network according to any one of claims 3 to 5, wherein the priority mobility is represented by the following formula:

In the formula, Q _j is the priority mobility rate of the LEO satellite switch node s _j ; λ _j is the data flow request rate of the migrated LEO satellite switch node s _j ; d _ij is the migration LEO satellite switch node s _j and the migrated LEO satellite controller node c The shortest distance between _i ; f _i is the processing capability of the LEO satellite controller c _i . 7. The satellite network multi-controller dynamic deployment method according to any one of claims 3 to 5, wherein the objective function is represented by the following formula: W _min =α×BL+β×newcost+γ×newT _a In the formula, W _min is the minimum value based on the migration overhead of the LEO satellite network, the post-migration LEO satellite control link delay, and the load balancing parameter of the post-migration LEO satellite network; BL is the load balancing parameter of the post-migration LEO satellite network; newcost is the normalized value of the LEO satellite network migration cost; newT _a is the normalized value of the LEO satellite control link delay after migration; α, β, and γ are the LEO satellite network migration cost, after migration The LEO satellite controls the link delay and different weights of the load balancing parameters of the LEO satellite network after migration, α+β+γ=1, 0≤α,β,γ≤1. 8. The method for dynamically deploying multiple controllers in a satellite network according to any one of claims 3 to 5, wherein the objective function is solved by using a whale optimization algorithm and a simulated annealing algorithm. 9. The method for dynamically deploying multiple controllers in a satellite network according to any one of claims 3 to 5, wherein the step S400 comprises: The migration LEO satellite controller node selects the LEO satellite switch node to be migrated, deploys migration rules to the LEO satellite switch node to be migrated, the LEO satellite switch node to be migrated sends a request to the target domain, and the target domain accepts the request of the LEO satellite switch node to be migrated , migrate the LEO satellite switch node to be migrated to the target domain; If the LEO satellite network is in an underload state, after migrating the LEO satellite switch node to be migrated to the target domain, shut down the underloaded LEO satellite controller node. 10. An SDN-based satellite network multi-controller dynamic deployment system, comprising a satellite network construction module, a load state determination module, a migration strategy formulation module, a migration module, an update module and a judgment module; The satellite network building module is used to construct an SDN-based satellite network architecture; The load state determination module is used to obtain the global load information of the LEO satellite network, and determine the load state of the LEO satellite network based on three thresholds; the load state of the LEO satellite network includes an overall overload state, a partial overload state, an underload state and a normal state. state; The migration strategy formulation module is configured to formulate a migration strategy based on the load state of the LEO satellite network when the LEO satellite network is in an overall overload state, a partial overload state or an underload state, and when the LEO satellite network is in a normal state In state, output LEO satellite network information; The migration module is configured to perform dynamic migration based on the migration strategy; The update module is used to update the mapping relationship between the LEO satellite switch node and the LEO satellite controller node in the LEO satellite network to obtain the current LEO satellite network; The judging module is used for judging the load state of the current LEO satellite network based on three thresholds, and if the current load state of the LEO satellite network is in a normal state, outputting the relevant information of the current LEO satellite network.

Images