A SDN controller deployment method for low-orbit satellite Internet of Things

Abstract

The present invention belongs to the field of satellite communication technology, and discloses a SDN controller deployment method for low-orbit satellite Internet of Things, including step 1 visibility analysis: using the expression of divided grid coordinates and low-orbit satellite coverage area, analyzing the number of low-orbit satellites covering a given grid, step 2 access mode selection: the satellite side estimates the access success rate of the satellite, and counts the number of terminals and loads that finally successfully access; the Internet of Things terminal side calculates the satisfaction, network activity factor and network state variables of access to all satellites covering the grid, analyzes the access probability of different satellites covering the grid, and selects the satellite to be accessed; step 3 controller deployment strategy optimization: establish a controller deployment optimization problem to determine the minimum number of controllers required and the corresponding deployment location of the controller. The present invention can meet the access load and end-to-end transmission delay of the low-orbit satellite Internet of Things by optimizing the number of SDN controls and deployment locations.

Description

Technical Field

[0001] The present invention relates to the field of satellite communication technology, and in particular to a method for deploying an SDN controller for a low-orbit satellite Internet of Things. Background Art

[0002] The application of the Internet of Things is gradually penetrating into various fields of human activities. However, due to the limitations of space and environment, the ground Internet of Things cannot provide reliable services for some cross-regional, harsh environment and other data collection fields, resulting in a mismatch between service capabilities and demand. The main reason is that for the Internet of Things that relies on wireless access, there must be a communication network built by a sufficient number of base stations. However, in some remote areas that cannot be reached by humans, it is very difficult to deploy base stations and establish communication networks: (1) Base stations cannot be built in large areas of oceans and deserts; (2) The cost of building base stations in remote areas with sparse populations is high; (3) Natural disasters can cause serious damage to ground networks. Therefore, how to establish communication networks in remote areas or areas that are difficult for people to reach is a difficult problem that must be overcome to realize the Internet of Everything. Among the many candidates, the low-orbit satellite communication system has become one of the reliable choices for assisting the ground network to realize the Internet of Everything because of its low transmission loss and delay characteristics, and the advantage of achieving seamless coverage of the earth through constellations.

[0003] However, with the increase in the number of satellites, a series of problems have arisen, such as the large number of satellite network nodes, rapid changes in topology, and greater difficulty in network management. A new network management method is needed. At this time, Software Defined Network (SDN) came into being, and SDN formed a new network architecture. The introduction of SDN further improves the scalability and flexibility of the integrated space-ground information network. In order to meet the low-latency and high-reliability service requirements of emergency tasks, multiple SDN controllers need to be deployed to achieve distributed control of satellite networks. Therefore, the number and location of controllers are key issues that need to be considered in designing controller placement plans and improving satellite network flexibility.

[0004] However, although the solution of placing controllers in the entire LEO layer based on satellite constellations can fully utilize the low latency characteristics of the LEO layer, it requires the selection of a large number of satellite nodes as controllers to meet the controller's associated coverage of switches. When the number of nodes increases and the dynamicity increases, the control nodes need to be synchronized to maintain a global network view. At this time, a large number of signaling interactions between nodes lead to increased network latency, which still cannot well meet the low latency requirements of dynamic networking for emergency tasks. Summary of the invention

[0005] In order to solve the above problems, the present invention provides an SDN controller deployment method for low-orbit satellite Internet of Things. Through the business model in the satellite Internet of Things scenario, the satellite business volume based on load balancing is obtained; by establishing the minimization of the number of controllers as the optimization goal and using a genetic algorithm for solving, the optimal number of controllers and placement scheme are determined; it can effectively find a suitable solution for multiple controllers in a short time, and ensure that the network has good performance in terms of latency and load balancing.

[0006] In order to achieve the above object, the present invention is implemented by the following technical solutions:

[0007] The present invention is a SDN controller deployment method for low-orbit satellite Internet of Things, including visibility analysis, access mode selection and controller deployment strategy optimization, and specifically includes the following steps:

[0008] Step 1: Visibility analysis: Based on the given longitude and latitude division accuracy, the earth's surface is evenly divided into multiple grids according to longitude and latitude, and for any grid, the conditions for low-orbit satellites to cover the grid are analyzed, and then the analyzed conditions are used to obtain the set of satellite numbers covering the grid, and the number of satellites in the set is counted to obtain the number of low-orbit satellites covering the grid;

[0009] Step 2: Access mode selection: The satellite side estimates the access success rate of the satellite based on the estimated load, and counts the number of terminals that successfully access and the load. The networked terminal side calculates the satisfaction, network activity factor and network state variable of all satellites covering the grid based on the number of active terminals in the grid and the elevation angle between the grid center and the visible satellite, combined with the available pilot number of the visible satellite and the estimated access success rate, and then analyzes the access probability of different satellites covering the grid, and selects the satellite to be accessed based on the analysis results;

[0010] Step 3: Optimize controller deployment strategy: Given the latency and access load, establish a controller deployment optimization problem to determine the minimum number of controllers required and the corresponding deployment locations of the controllers.

[0011] A further improvement of the present invention is that in step 1, given a grid g, the condition for a low-orbit satellite n to completely cover the grid is:

[0012]

[0013] Among them, r b represents the radius of the Earth, represents the satellite elevation angle, r h represents the satellite altitude, r z =r b +r h , the maximum distance from the satellite to the coverage area (x l ,y l ,z l ) T represents the vertex coordinates of mesh g, and l represents the vertex of mesh g.

[0014] A further improvement of the present invention is that step 2 specifically comprises the following steps:

[0015] Step 2-1: Given a grid g and the terminal deployment density λ of the grid g , calculate the number of new services generated in each time slot, assuming that the IoT terminal is between t = 0 and T S If a certain time node between is triggered, the expression of the number of terminals triggered in the grid g in the i-th time slot is:

[0016]

[0017] Where: S g represents the area of ​​the grid g, λ g represents the terminal density of the grid g, N g is the total number of grids, 1≤g≤N g , p(t) represents the probability distribution function of Beta distribution, α and β are Beta distribution parameters, α>0, β>0, t i =it s represents the time corresponding to the i-th time slot, t s Indicates the time slot length, T S Indicates the total time, go to step 2-2;

[0018] Step 2-2: Satellite n estimates the load, i.e. the number of terminals connected, based on its own preamble usage And calculate the access success rate, The expression is:

[0019]

[0020] Where: I i-1,n is the number of idle preamble sequences of satellite n in the i-1th time slot, R i-1,n is the number of preamble sequences available to satellite n, M i-1,n is the number of payloads that successfully access satellite n in the i-1th time slot, lambertw(·) represents the Lambert w function, and the access success rate of satellite n in the i-th time slot is:

[0021]

[0022] Where: M i,n is the number of terminals successfully accessing satellite n in the i-th time slot, Estimate the total number of terminals accessing satellite n in the i-th time slot, and proceed to step 2-3;

[0023] Step 2-3: Calculate the satellite elevation angle based on the center coordinates of the grid g where the terminal is located and the real-time sub-satellite point of the satellite Combined with the number of satellite coverage grids q i,n , access success rate P i,n and the available leading derivative R i,n Calculate the satisfaction of satellite n Thus, the network state variable of satellite n is obtained Thus, the probability of access for each visible satellite is obtained Go to step 2-4;

[0024] Step 2-4: For grid g, the IoT terminals activated in the i-th time slot generate random numbers R that obey a uniform distribution h ~U(0,1), and the cumulative probability matrix The elements in the matrix are compared in turn, and the first element number in the matrix that is greater than the random number is taken as the satellite number of each random access. g Indicates the number of satellites covering grid g, and proceeds to step 2-5;

[0025] Step 2-5: The terminal generates a random number μ between 0 and 1, and the satellite n broadcasts the access level restriction parameter λ n , if μ≤λ, the terminal needs to access, otherwise, back off and retransmit; where λ n The expression is:

[0026]

[0027] Where: R in is the number of preamble sequences available to satellite n, Estimate the number of terminals connected to the satellite and proceed to step 2-6;

[0028] Step 2-6: After the number of terminals is controlled, random access requests are sent to the corresponding satellites. If a collision occurs during this process, back off and resend; if no collision occurs, the number of terminals that successfully access and the number of idle preamble resources are counted.

[0029] A further improvement of the present invention is that in step 3, the average total delay expression of the end-to-end transmission of the network is:

[0030]

[0031] Where: V = {v1,v2,...,v M} represents all independent switch nodes, M is the total number of switches, C = {c1, c2, ..., c K} represents all controller nodes, K is the number of controllers, x cv Assign the switch matrix X = [x cv ] K×M The elements in the satellite have only two values, 0 and 1, which represent the control relationship between the controller and the switch. d represents the shortest distance between any two nodes in the satellite. w (,)The shortest distance between any two nodes in the satellite network, w∈{1,2}.

[0032] The static load balancing rate expression of the controller is:

[0033]

[0034] Where, K is the number of controllers, and the controller load value 1≤c≤K, switch V v Load value 1≤v≤M, average load of controller

[0035] The optimization model expression based on minimizing the number of controllers is:

[0036]

[0037] st

[0038]

[0039]

[0040] T total ≤T

[0041] B rate ≤B

[0042] Where, K is the number of controllers, x cv Assign the switch matrix X = [x cv ] K×M The elements in V can only have two values: 0 and 1. M} represents all independent switch nodes, C = {c1, c2, ..., c K} represents all controller nodes, B rate is the static load balancing rate of the controller, T total is the average total delay of the end-to-end transmission of the network, T is the maximum average total delay allowed by the network, B is the maximum load imbalance rate of the controller allowed by the network, N is the number of visible satellites, and x cv Indicates the control relationship between the switch and the controller.

[0043] A further improvement of the present invention is that step 3 specifically comprises the following steps:

[0044] Step 3-1: Get the satellite node distance d w , the number of visible satellites N and satellite number O N , the traffic volume of each satellite q n , 1≤n≤N, randomly generate L chromosomes to form a chromosome population chrom, where each chromosome represents a controller deployment method;

[0045] Step 3-2: Calculate the fitness of the chromosome population, the expression is:

[0046] Fit=ω1T total +ω2B rate

[0047] Among them: ω1, ω2 are weight coefficients; B rate is the static load balancing rate of the controller, T total is the average total delay of end-to-end transmission in the network;

[0048] Step 3-3: Perform multiple different selection, crossover, and mutation operations on the chromosomes in the population to obtain a new chromosome population newchrom;

[0049] Step 3-4: Calculate the fitness of the new chromosome population, and perform multiple different selection, crossover, and mutation operations on the chromosomes in the population to obtain a new chromosome population newchrom;

[0050] Step 3-5: Iterate steps 3-3 and 3-4. When the number of iterations exceeds the algorithm termination condition, terminate the loop and output the number of controllers and deployment plan.

[0051] The beneficial effects of the present invention are as follows: the present invention targets the characteristics of dynamic changes in satellite networks, combines the service model under the low-orbit satellite Internet of Things, and obtains satellite service volume based on load balancing;

[0052] The present invention establishes an optimization model for minimizing the number of controllers and uses a genetic algorithm to solve it, which can not only find the best controller deployment strategy in a short time, but also ensure that the network has good performance in terms of latency and load balancing.

[0053] The present invention can meet the access load and end-to-end transmission delay of low-orbit satellite Internet of Things by optimizing the number of SDN controls and the deployment location. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a flowchart of the implementation process of the method of the present invention.

[0055] Figure 2 The figure is a comparison diagram of the relationship between two different algorithms and two different load balancing rates in the method of the present invention.

[0056] Figure 3 The figure is a comparison diagram of the relationship between two different algorithms and two different network delays in the method of the present invention. DETAILED DESCRIPTION

[0057] The following will disclose the embodiments of the present invention with drawings. For the purpose of clear description, many practical details will be described together in the following description. However, it should be understood that these practical details should not be used to limit the present invention. That is to say, in some embodiments of the present invention, these practical details are not necessary.

[0058] like Figure 1 The present invention is an SDN controller deployment method for low-orbit satellite Internet of Things. The SDN controller deployment method includes three parts: visibility analysis, access mode selection, and controller deployment strategy optimization, and specifically includes the following steps:

[0059] Step 1: Visibility Analysis:

[0060] Based on the given longitude and latitude division accuracy, the earth's surface is evenly divided into multiple grids according to longitude and latitude, and for any grid, the conditions for low-orbit satellites to cover the grid are analyzed, and then the analyzed conditions are used to obtain the set of satellite numbers covering the grid, and the number of satellites in the set is counted to obtain the number of low-orbit satellites covering the grid;

[0061] Given a grid g, the condition for a low-orbit satellite n to completely cover the grid is:

[0062]

[0063] The radius of the Earth is r b , satellite elevation angle Satellite altitude r h , r z =r b +r h , the maximum distance from the satellite to the coverage area (x l ,y l ,z l ) T represents the vertex coordinates of mesh g, and l represents the vertex of mesh g.

[0064] Step 2: Select the access method:

[0065] The satellite side estimates the possible access success rate of the satellite based on the estimated load, and counts the number of terminals that finally successfully access and the load; the IoT terminal side calculates the satisfaction level, network activity factor and network state variables of all satellites covering the grid based on the number of active terminals in the grid and the elevation angle between the grid center and the visible satellite, combined with the available pilot number of the visible satellite and the estimated access success rate, and then analyzes the access probability of different satellites covering the grid, and selects the satellite to be accessed based on the analysis results.

[0066] (2.1) A certain area on the surface of the earth is divided into grids with the division accuracy of Δlat and Δlon respectively. Δlat means that the latitude accuracy is 2°, and Δlon means that the longitude accuracy is 2.5°. This means that within the area, a grid is divided every 2° in latitude and every 2.5° in longitude. Business analysis is performed against the terminal distribution density in the satellite IoT scenario to determine the terminal deployment density λ within a given range. g , go to step (2.2);

[0067] (2.2) Given a grid g and the terminal deployment density λ of the grid g , calculate the number of new services generated in each time slot, using the Beta (α, β) distribution, assuming that the IoT terminal is between t = 0 and T s If a certain time node between is triggered, the expression of the number of terminals triggered in the grid g in the i-th time slot is:

[0068]

[0069] Where: S g represents the area of ​​the grid g, λ g represents the terminal density of the grid g, N g is the total number of grids, 1≤g≤N g , p(t) represents the probability distribution function of Beta distribution, α and β are Beta distribution parameters. In the embodiment of the present invention, α=3, β=4. t i =it s represents the time corresponding to the i-th time slot, t s Indicates the time slot length, T S Indicates the total time, go to step (2.3);

[0070] (2.3) Satellite n estimates the number of terminals that access based on its own preamble usage And calculate the access success rate. The satellite side load estimation expression at this time is:

[0071]

[0072] Where: Ii-1,n is the number of idle preamble sequences of satellite n in the i-1th time slot, R i-1,n is the number of preamble sequences available to satellite n, M i-1,n is the number of payloads that successfully access satellite n in the i-1th time slot, and lambertw(·) represents the Lambert w function.

[0073] In the i-th time slot, the possible access success rate of satellite n is:

[0074]

[0075] Where: M i,n is the number of terminals successfully accessing satellite n in the i-th time slot, Estimate the total number of terminals accessed by satellite n in the i-th time slot and proceed to step (2.4);

[0076] (2.4) Calculate the satellite elevation angle based on the center coordinates of the grid g where the terminal is located and the real-time satellite sub-satellite point Combined with the number of satellite coverage grids q i,n , access success rate P i,n and the available leading derivative R i,n Calculate the satisfaction of satellite n Thus, the network state variable of satellite n is obtained Thus, the probability of access for each visible satellite is obtained Go to step (2.5);

[0077] (2.5) For grid g, the IoT terminals activated in the i-th time slot generate random numbers R that follow a uniform distribution h ~U(0,1), and the cumulative probability matrix The elements in the matrix are compared in turn, and the first element number in the matrix that is greater than the random number is taken as the satellite number of each random access. g represents the number of satellites covering grid g, and proceed to step (2.6);

[0078] (2.6) The terminal generates a random number μ between 0 and 1, and the satellite n broadcasts the access level restriction parameter λ n , if μ≤λ, the terminal needs to access, otherwise, back off and retransmit; where λ n The expression is:

[0079]

[0080] Where: R i,n is the number of preamble sequences available to satellite n, Estimate the number of terminals connected to the satellite and proceed to step (2.7);

[0081] (2.7) After the number of terminals is controlled, random access requests are sent to the corresponding satellites. This process may cause collisions. If a collision occurs, back off and resend. If there is no collision, the number of terminals that successfully access and the number of idle preamble resources are counted.

[0082] Step 3: Optimize controller deployment strategy:

[0083] When the latency and access load are given, a controller deployment optimization problem is established to determine the minimum number of controllers required and the corresponding deployment locations of the controllers.

[0084] The average total delay expression of end-to-end transmission in the network is:

[0085]

[0086] Where: V = {v1,v2,...,v M} represents all independent switch nodes, M is the total number of switches, C = {c1, c2, ..., c K} represents all controller nodes, K is the total number of controllers to be deployed, x cv Assign the switch matrix X = [x cv ] K×M The elements in have only two values: 0 and 1, representing the control relationship between the controller and the switch.

[0087] Divide the period into T a Time slice, each time node of the change is {t1,t2,...,t Ta},sat i ,sat j are satellite nodes. Any two nodes in the satellite network are i and sat j The shortest distance between where w∈{1,2}, sat i ∈{C,V}, Indicates that at t z Slot Node sat i and sat j The shortest distance between.

[0088] The static load balancing rate expression of the controller is:

[0089]

[0090] Where, K is the number of controllers, and the controller load value 1≤c≤K, switch V v Load value 1≤v≤M, average controller load

[0091] The optimization model expression based on minimizing the number of controllers is:

[0092]

[0093] st

[0094]

[0095]

[0096] T total ≤T

[0097] B rate ≤B

[0098] The specific steps of the algorithm are as follows:

[0099] (3.1) Get the satellite node distance d w , the number of visible satellites N and satellite number O N , the traffic volume of each satellite q n , 1≤n≤N, randomly generate L chromosomes to form a chromosome population chrom, where each chromosome represents a controller deployment method;

[0100] (3.2) Calculate the fitness of the chromosome population, the expression is:

[0101] Fit=ω1T total +ω2B rate

[0102] Wherein: ω1, ω2 are weight coefficients. In the embodiment of the present invention, ω1=0.9, ω2=0.1 is in line with the actual situation. total is the average total end-to-end delay of the current chromosome, B rate is the static load balancing rate of the controller in the current chromosome;

[0103] (3.3) Perform multiple different selection, crossover, and mutation operations on the chromosomes in the population to obtain a new chromosome population newchrom;

[0104] (3.4) Calculate the fitness of the new chromosome population, and perform multiple different selection, crossover, and mutation operations on the chromosomes in the population to obtain a new chromosome population newchrom;

[0105] (3.5) Loop through steps (3.3)-(3.4). When the number of iterations exceeds the algorithm termination condition, terminate the loop and output the number of controllers and deployment plan.

[0106] The present invention aims at the dynamic characteristics of satellite networks, combines the business model under the low-orbit satellite Internet of Things, and obtains the satellite business volume based on load balancing; establishes an optimization model that minimizes the number of controllers, and uses a genetic algorithm to solve it, which can find a suitable deployment plan in a short time and effectively improve the performance of the overall network in terms of latency and load balancing.

[0107] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for deploying an SDN controller for low-orbit satellite Internet of Things, characterized by: The SDN controller deployment method includes three parts: visibility analysis, access mode selection and controller deployment strategy optimization, and specifically includes the following steps: Step 1: Visibility analysis: Based on the given longitude and latitude division accuracy, the earth's surface is evenly divided into multiple grids according to longitude and latitude, and for any grid, the conditions for low-orbit satellites to cover the grid are analyzed, and then the analyzed conditions are used to obtain the set of satellite numbers covering the grid, and the number of satellites in the set is counted to obtain the number of low-orbit satellites covering the grid; Step 2: Access mode selection: The satellite side estimates the access success rate of the satellite based on the estimated load, and counts the number of terminals that successfully access and the load. The networked terminal side calculates the satisfaction, network activity factor and network state variable of all satellites covering the grid based on the number of active terminals in the grid and the elevation angle between the grid center and the visible satellite, combined with the available pilot number of the visible satellite and the estimated access success rate, and then analyzes the access probability of different satellites covering the grid, and selects the satellite to be accessed based on the analysis results; Step 3: Optimizing controller deployment strategy: When the latency and access load are given, a controller deployment optimization problem is established to determine the minimum number of controllers required and the corresponding deployment locations of the controllers. In step 3, the optimization model expression based on minimizing the number of controllers is: T total ≤T B rate ≤B Where, K is the number of controllers, x cv Assign the switch matrix X = [x cv ] K×M The elements in V can only have two values: 0 and 1. M } represents all independent switch nodes, C = {c1, c2, ..., c K } represents all controller nodes, B rate is the static load balancing rate of the controller, T total is the average total delay of the end-to-end transmission of the network, T is the maximum average total delay allowed by the network, B is the maximum load imbalance rate of the controller allowed by the network, N is the number of visible satellites, and x cv Indicates the control relationship between the switch and the controller; The step 3 specifically includes the following steps: Step 3-1: Get the satellite node distance d w , the number of visible satellites N and satellite number O N , the traffic volume of each satellite q n , 1≤n≤N, randomly generate L chromosomes to form a chromosome population chrom, where each chromosome represents a controller deployment method; Step 3-2: Calculate the fitness of the chromosome population, the expression is: Fit=ω1T total +ω2B rate Among them: ω1, ω2 are weight coefficients; B rate is the static load balancing rate of the controller, T total is the average total delay of end-to-end transmission in the network; Step 3-3: Perform multiple different selection, crossover, and mutation operations on the chromosomes in the population to obtain a new chromosome population newchrom; Step 3-4: Calculate the fitness of the new chromosome population, and perform multiple different selection, crossover, and mutation operations on the chromosomes in the population to obtain a new chromosome population newchrom; Step 3-5: Iterate steps 3-3 and 3-4. When the number of iterations exceeds the algorithm termination condition, terminate the loop and output the number of controllers and deployment plan.

2. According to claim 1, a SDN controller deployment method for low-orbit satellite Internet of Things is characterized by: In step 1, given a grid g, the condition for a low-orbit satellite n to completely cover the grid is: Among them, r b represents the radius of the Earth, represents the satellite elevation angle, r h represents the satellite altitude, r z =r b +r h , the maximum distance from the satellite to the coverage area (x l ,y l ,z l ) Τ represents the vertex coordinates of mesh g, and l represents the vertex of mesh g.

3. According to claim 1, a SDN controller deployment method for low-orbit satellite Internet of Things is characterized by: The step 2 specifically includes the following steps: Step 2-1: Given a grid g and the terminal deployment density λ of the grid g , calculate the number of new services generated in each time slot, assuming that the IoT terminal is between t = 0 and T S If a certain time node between is triggered, the expression of the number of terminals triggered in the grid g in the i-th time slot is: Where: S g represents the area of ​​the grid g, λ g represents the terminal density of the grid g, N g is the total number of grids, 1≤g≤N g , p(t) represents the probability distribution function of Beta distribution, α and β are Beta distribution parameters, α>0, β>0, t i =it s represents the time corresponding to the i-th time slot, t s Indicates the time slot length, T S Indicates the total time, go to step 2-2; Step 2-2: Satellite n estimates the load, i.e. the number of terminals connected, based on its own preamble usage And calculate the access success rate, The expression is: Where: I i-1,n is the number of idle preamble sequences of satellite n in the i-1th time slot, R i-1,n is the number of preamble sequences available to satellite n, M i-1,n is the number of payloads that successfully access satellite n in the i-1th time slot, lambertw(·) represents the Lambert w function, and the access success rate of satellite n in the i-th time slot is: Where: M i,n is the number of terminals successfully accessing satellite n in the i-th time slot, Estimate the total number of terminals accessing satellite n in the i-th time slot, and proceed to step 2-3; Step 2-3: Calculate the satellite elevation angle based on the center coordinates of the grid g where the terminal is located and the real-time sub-satellite point of the satellite Combined with the number of satellite coverage grids q i,n , access success rate P i,n and the available leading derivative R i,n Calculate the satisfaction of satellite n Thus, the network state variable of satellite n is obtained Thus, the probability of access for each visible satellite is obtained Go to step 2-4; Step 2-4: For grid g, the IoT terminals activated in the i-th time slot generate random numbers R that obey a uniform distribution h ~U(0,1), and the cumulative probability matrix The elements in the matrix are compared in turn, and the first element number in the matrix that is greater than the random number is taken as the satellite number of each random access. g Indicates the number of satellites covering grid g, and proceeds to step 2-5; Step 2-5: The terminal generates a random number μ between 0 and 1, and the satellite n broadcasts the access level restriction parameter λ n , if μ≤λ, the terminal needs to access, otherwise, back off and retransmit; where λ n The expression is: Where: R i,n is the number of preamble sequences available to satellite n, Estimate the number of terminals connected to the satellite and proceed to step 2-6; Step 2-6: After the number of terminals is controlled, random access requests are sent to the corresponding satellites. If a collision occurs during this process, back off and resend; if no collision occurs, the number of terminals that successfully access and the number of idle preamble resources are counted.

4. According to claim 1, a SDN controller deployment method for low-orbit satellite Internet of Things is characterized by: In step 3, the average total delay expression of the end-to-end transmission of the network is: Where: V = {v1,v2,...,v M } represents all independent switch nodes, M is the total number of switches, C = {c1, c2, ..., c K } represents all controller nodes, K is the number of controllers, x cv Assign the switch matrix X = [x cv ] K×M The elements in the satellite have only two values, 0 and 1, which represent the control relationship between the controller and the switch. d represents the shortest distance between any two nodes in the satellite. w (,)The shortest distance between any two nodes in the satellite network, w∈{1,2}; The static load balancing rate expression of the controller is: Where, K is the number of controllers, and the controller load value Switch V v Load value T a is the time slice, 1≤v≤M, the average load of the controller

Images