Novel low-power-consumption lossy network routing method

Abstract

The invention discloses a novel low-power-consumption lossy network routing method, which comprises the following steps of: according to a membership function of each routing metric, constructing a comprehensive membership function containing each routing metric on the basis of a triangular module fusion operator; designing a context awareness comprehensive objective function based on the comprehensive membership function and the maximum membership principle; designing a calculation mechanism of a node rank based on a context awareness comprehensive objective function, and selecting an optimal path according to a calculation mechanism of a new next hop node; the routing method is applied to a low-power-consumption lossy network environment. According to the invention, wide application of a low-power-consumption lossy network routing protocol is promoted.

Description

Technical Field

[0001] The present invention relates to the field of low power consumption, and in particular to a new low power consumption lossy network routing method. Background Art

[0002] The Routing Protocol for Low-power and Lossy Networks (RPL) is a distance vector routing protocol specially developed by IETF (Internet Engineering Task Force) for low-power and lossy networks. Routing Metric (RM) is an important basis for constructing objective functions, calculating path costs, and selecting optimal paths in routing protocols. RFC6551 published by IETF details the routing metrics and calculation methods that can be used by RPL. Comprehensively evaluating multiple routing metrics based on different application scenario requirements can effectively improve network life, network topology stability, packet delivery success rate, latency and other performance.

[0003] However, the multi-routing metric evaluation method has not been effectively applied and promoted, mainly because there is no systematic and scientific multi-routing metric evaluation method theory to rely on.

[0004] Most of the existing RPL multi-routing metric evaluation algorithms are based on the lexical method (Lexical) or linear weighted method (Additive) to construct the objective function. Among them, the effectiveness of the lexical method is generally lower than that of the linear weighted method, but the determination of the weights of each routing metric in the linear weighted method is mostly based on the subjective personal experience of experts, lacking a scientific and reasonable weight distribution theory, which to a certain extent restricts the development and widespread application of multi-routing metric evaluation methods. For example: The literature [ZQ.Wang, ZK.Jin, et al. Increasing efficiency for routing in internet of things usingBinary Gray Wolf Optimization and fuzzy logic[J].JOURNAL OF KING SAUDUNIVERSITY-COMPUTER AND INFORMATION SCIENCES, 2023, 35(9): 1-17.] proposes to use the gray wolf algorithm and fuzzy logic method to evaluate node spacing and residual energy, but the fuzzy logic method simply fuzzifies the information, which will reduce the control accuracy and dynamic quality of the system. The literature [J.Mohajerani, MM.Ghanatghestani, et al.MCTE-RPL:A multi-context trust-based efficient RPL for IoT[J].JOURNAL OF NETWORKAND COMPUTER APPLICATIONS, 2024, 229] proposed to use a linear weighted method to evaluate the remaining energy, reliable ETX and node rank value, but did not propose a scientific and reasonable basis for the allocation of routing metric weights, and could not effectively improve network performance.

[0005] In summary, the existing RPL multi-routing metric evaluation algorithm has the following main problems:

[0006] (1) Routing metric evaluation is not comprehensive

[0007] Most studies evaluate ETX and node residual energy, but do not consider other routing metrics such as node cache occupancy. As a result, it is impossible to take into account the performance of load balancing, energy consumption, latency, stability, etc., which restricts the promotion and application of the protocol.

[0008] (2) Traditional calculation methods of ETX and latency affect network performance to a certain extent

[0009] The traditional ETX calculation method fails to consider the problem that there may be one or more links with large ETX on the path with the smallest ETX value, and this link may be the bottleneck of the entire network; similarly, the calculation of delay metrics also has the above problem.

[0010] (3) The design of the objective function and the weight allocation method of each routing metric are unreasonable

[0011] The objective function is mainly used to specify how to combine several routing metrics into a composite metric and convert it into a node rank value, based on which the optimal path is selected. The objective function can design different optimal path selection rules according to the actual network application scenario requirements. The weight of the routing metric indicates the importance of the metric. At present, the determination of the weight of each routing metric is mostly based on the subjective experience of experts. This method is too subjective and cannot take into account the objective actual operation of the network at the same time, which affects the performance of the network.

[0012] (4) The next-hop node selection mechanism is too simple and crude

[0013] Failure to sort the positive and negative ideal solutions of each neighbor node will affect the selection result of the next-hop node to a certain extent, thereby affecting the network performance. Summary of the invention

[0014] The present invention provides a new low-power lossy network routing method. The present invention selects the comprehensive optimal path to transmit data, thereby optimizing network life, reliability and other aspects of performance; and provides a certain scientific theoretical basis for multi-route metric evaluation, promoting the widespread application of low-power lossy network routing protocols, as described below:

[0015] A new low-power lossy network routing method, situation-aware path conditions, and scientific evaluation of multiple routing metrics, the method includes:

[0016] According to the membership functions of each routing metric, a comprehensive membership function including each routing metric is constructed based on the triangle module fusion operator;

[0017] Design the situational awareness comprehensive objective function based on the comprehensive membership function and the maximum membership principle;

[0018] Design a node rank calculation mechanism based on the context-aware comprehensive objective function, and select the optimal path based on the calculation mechanism of the new next-hop node;

[0019] Use the best path for low-power lossy network environments.

[0020] Wherein, the comprehensive membership function is:

[0021]

[0022] In the formula, i represents the i-th neighbor node, g j (i) represents the membership function of each routing metric of neighbor node i.

[0023] Wherein, the situational awareness comprehensive objective function is:

[0024] OF CA =min(OF CA (i))

[0025]

[0026] The calculation mechanism of the node rank is:

[0027] 1) Node rank calculation design

[0028] The rank value of the root node is set to 1.0, and the ranks of other non-root nodes are calculated. c is a common non-root node, i is the neighbor of c, and R c (i) indicates the rank value corresponding to node c when node c selects i as the next hop, R cp (i) is the rank value of node i. If c has n neighbor nodes, find {R c (1),R c (2),…,R c (n)}; if min{R c (1),R c (2),…,R c (n)}=R c (f), select node f as the next hop of c;

[0029] R c (i) = R cp (i)+(OF CA (i)+1),i=1,2,…,n

[0030] 2) Design of next-hop node selection mechanism

[0031] (1) If the rank calculated by a neighbor node is smaller than the rank calculated by the current next-hop node, but the difference between the two is smaller than the next-hop node change threshold, the current next-hop node will continue to be used to transmit data without changing the next-hop node;

[0032] (2) If the calculated rank value is less than or equal to 1 or greater than the total number of nodes in the network, it means that the rank value calculated through the neighbor node is wrong or the path quality is extremely poor or the node processing capacity on the path is extremely low, and it should be recalculated; if the result is still the same, the node is directly set as a leaf node;

[0033] (3) If there are multiple neighbor nodes at the same time, and the rank value of node c obtained through their paths is the smallest and equal, then the node with the largest set of neighbor nodes is selected as the preferred parent node;

[0034] (4) If node c has only one neighbor, node c needs to wait for a while to obtain more neighbor nodes. If the number of neighbor nodes of node c is greater than or equal to 2, node c selects the next hop node by executing the CR-T method. If the number of neighbors of node c is still 1, node c directly selects the neighbor as the next hop, and the rank of node c is equal to the rank value of the neighbor plus 1.

[0035] The beneficial effects of the technical solution provided by the present invention are:

[0036] (1) Save energy and improve network life

[0037] ① When node c has only one neighbor, it can directly select the neighbor as the next hop node without executing any protocol process, saving the transmission of control information, saving energy consumption, and thus extending the network life.

[0038] ② When the difference between the rank calculated by node c through a neighbor and the current rank is less than the threshold for changing the next-hop node, the next-hop node remains unchanged. This will also save energy to a certain extent and extend the network life.

[0039] ③ Through situational awareness, not only the residual energy index of the neighboring node is evaluated, but also the residual energy index of the next hop of the neighboring node is evaluated, avoiding the selection of the path with lower residual energy to transmit data on the entire path, thereby extending the network life.

[0040] (2) Optimized network reliability and latency performance

[0041] CR-T proposes an improved ETX and delay calculation method to ensure that the selected path has low ETX and delay and avoid the situation where a single link has a large ETX or delay, thereby optimizing the reliability and delay performance of the network.

[0042] (3) Load balancing to avoid network congestion

[0043] Through situational awareness, not only the cache usage of neighboring nodes is evaluated, but also the cache usage of the next hop of the neighboring node is evaluated, avoiding the selection of congested paths to transmit data along the entire path, thereby achieving network load balancing.

[0044] In addition, by designing new methods such as the triangle mode fusion operator, the context-aware comprehensive objective function, and the node rank calculation mechanism, we can comprehensively evaluate the relevant important routing metrics and then find the optimal path to transmit data. This can improve the network performance in terms of network life, latency, reliability, etc.

[0045] (4) The present invention can be applied to low-power lossy network environments, such as smart grid neighborhood networks. The neighborhood network is mainly responsible for the interconnection between the home area network and the wide area network, and is composed of multiple smart terminals and aggregation nodes. The present invention can be preferably applied to the neighborhood network. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 This is an example diagram of a low-power lossy network topology;

[0047] Figure 2 A simplified diagram of the construction process for DODAG;

[0048] Figure 3 Flow chart of the CR-T method. DETAILED DESCRIPTION

[0049] In order to make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention are described in further detail below.

[0050] Basic settings of the embodiments of the present invention

[0051] (1) Each node has a unique ID number. During the network initialization phase, the root node assigns network addresses to ordinary nodes joining the network and forms a network;

[0052] (2) Bidirectional data transmission can be carried out between any two nodes in the network;

[0053] (3) The communication range of the node is about 100 meters, and the node can be static or dynamic.

[0054] Specific implementation of the embodiment of the present invention

[0055] The specific implementation of the present invention is described as follows: there is a root node and multiple common nodes in the network. Bidirectional data transmission can be performed between any two nodes in the network. It includes three stages: network initialization, specific execution of the CR-T method and data transmission.

[0056] Low power lossy network topology model such as Figure 1As shown in the figure. The model contains two types of nodes: root nodes and ordinary nodes. The nodes in low-power and lossy networks (LLNs) are limited in storage capacity, energy consumption and processing power, and the links between nodes are unstable. It can be widely used in different fields such as urban networks and smart grids. The construction and maintenance of its network topology are mainly carried out through the destination-oriented directed acyclic graph (DODAG). All paths in the graph point to the DODAG root node, and there is no loop. DODAG is constructed and maintained through control messages (the explanation of the relevant control messages is shown in Table 1) and objective functions (OF).

[0057] Table 1 Control messages and DODAG explanation

[0058]

[0059] The construction and maintenance process of DODAG is as follows Figure 2 As shown, the specific construction and maintenance process is as follows:

[0060] Step 1: The root node broadcasts a DIO message to its neighbor nodes. The DIO contains the relevant information needed to build and maintain the DODAG.

[0061] Step 2: After receiving the DIO message, the neighboring node of the root node, such as node 1, determines whether to select the root node as its next hop node according to OF. If node 1 selects the root node as its next hop node, node 1 unicasts the DAO message to the root node to build a complete upstream and downstream route.

[0062] Step 3: Node 1 broadcasts the DIO message to its neighboring nodes, such as Node 2, again, and Node 2 performs the same operation as Node 1 in Step 2.

[0063] Step 4: After receiving DAO2 from node 2, node 1 unicasts DAO containing updated routing information 12 Message to the root node.

[0064] Step 5: Other nodes such as node 3 perform the same operations as the above steps to join the DODAG.

[0065] Step 6: If node 4 does not receive DIO messages from other nodes within a period of time, node 4 broadcasts a DIS message to its neighbor nodes, requesting its neighbor node (node ​​3) to send a DIO message to itself to join the DODAG.

[0066] Step 7: After node 4 joins the DODAG, node 3 unicasts the DAO with updated routing information to node 2 34Message, then node 2 unicasts the DAO with updated routing information to node 1 234 Message, finally node 1 unicasts the DAO with updated routing information to the root node 1234 information.

[0067] Repeat the above steps until the network covers all nodes.

[0068] Figure 3 This is a specific flowchart for the execution of the CR-T method. The relevant contents of the CR-T method will be introduced in detail later based on this flowchart.

[0069] 1. Membership Function and Triangle Module Fusion Operator

[0070] 1. Membership function

[0071] Definition 1: Let A be a mapping from the domain X to [0,1], that is, A:X→[0,1], x→A(x), then A is called a fuzzy set on X, and the function A(·) is called the membership function of the fuzzy set A, and A(x) is called the membership of x to the fuzzy set.

[0072] According to Definition 1, the membership function can be further defined as:

[0073] Definition 2: If there is a number F(x) (F(x)∈[0,1]) corresponding to any element x in the domain (research scope) U, then F is called a fuzzy set on U, and F(x) is called the membership of x to F. When x changes in U, F(x) is a function, called the membership function of x. The closer the membership function F(x) is to 1, the higher the degree to which x belongs to F, and the closer F(x) is to 0, the lower the degree to which x belongs to F.

[0074] The membership function F(x) with a value in the interval [0, 1] is used to represent the degree to which x belongs to F. This is a more reasonable way to describe fuzzy problems than classical set theory. The new CR-T method will use the membership function to represent the performance of each node's routing metric.

[0075] 2. Triangle modulus fusion operator

[0076] In order to obtain a comprehensive evaluation of the performance of various aspects of a system, the membership function of each evaluation index of the triangle mode fusion operator system can be used to achieve the advantage balance and contradiction neutralization of the influence of each metric index. That is, the evaluation result of a system is no longer simply good or bad, but a fuzzy set. For example, by using the triangle mode fusion operator, CR-T can efficiently fuse the membership functions of each routing metric of the neighboring nodes to comprehensively evaluate the neighboring nodes. Select the node with small ETX, small delay, and high residual energy as the next hop node, and exclude the neighboring nodes with poor routing metric values. The definition of the triangle mode fusion operator is as follows:

[0077] Definition 3: The mapping T:[0,1] 2 →[0,1] is the triangle norm (triangle norm operator or fuzzy operator), if The following conditions are met:

[0078] (1) T(0,0)=0, T(1,1)=1;

[0079] (2)T(a,b)≤T(c,d), if a≤c,b≤d;

[0080] (3) T(a,b)=T(b,a);

[0081] (4)T(T(a,b),c)=T(a,T(b,c)).

[0082] In addition, for practical applications, the triangle mode fusion operator can be extended to multiple dimensions. CR-T is to extend the triangle mode fusion operator to four dimensions.

[0083] 2. Context-Aware Routing Metrics and Their Membership Functions

[0084] 1. Context-aware routing metrics

[0085] In order to comprehensively evaluate the capabilities of neighboring nodes in all aspects, the embodiment of the present invention will comprehensively evaluate the following five routing metrics at the same time: node residual energy index REI (REI), buffer occupancy ratio BOR (BOR, Buffer Occupancy Ratio), ETX, delay D (Delay) and hop count (HC, Hop Count). Assuming that the non-root node c has n neighboring nodes, the standardization of each routing metric is as follows:

[0086] (1) Node Residual Energy Index (REI)

[0087] As shown in formula (1), REI represents the residual energy index of the neighboring node. initial (i) represents the maximum initial energy of neighbor node i, E current (i) represents the current remaining energy of neighbor node i, i p represents the next hop node of i, and β = 0.21 is an adjustment parameter used to adjust the influence of the next hop node of i. REI(i) reflects the influence of neighbor node i and its next hop node i. p The residual energy index of the node. It can be seen that REI(i) represents the residual energy of the node in an iterative manner, and the influence of the residual energy of the neighboring node decreases as the path goes deeper. This method of uplink and downlink situation-aware node residual volume can effectively avoid selecting a path with lower energy to transmit data.

[0088]

[0089] (2) Buffer Occupancy Rate (BOR)

[0090] BOR represents the cache usage of a node, and its calculation method is shown in formula (2). Where Q(i) is the cache usage index, which can be calculated according to formula (3). p represents the next hop node of neighbor node i, β = 0.21 is an adjustment parameter used to adjust the influence of the next hop node of i. BOR(i) reflects the influence of neighbor node i and its next hop node i. p It can be seen that BOR(i) represents the cache occupancy of the node in an iterative manner, and the impact of the cache occupancy of the neighboring node decreases as the path goes deeper. This way of sensing the node cache status in the uplink and downlink scenarios can effectively alleviate congestion and balance the load.

[0091]

[0092] (3)ETX

[0093] According to the traditional ETX calculation method, the ETX of a path is equal to the sum of the ETX of all links on the path. The path corresponding to the minimum ETX is the optimal path. This optimal path selection method may select the path with a larger ETX link as the optimal path. For this reason, the embodiment of the present invention proposes a method of combining the mean square error, sum and mean of the ETX of each link on the path. The path P from node c to the destination node via neighbor node i is i ETX sum (ETX(i)), ETX mean and ETX mean square error (σ ETX (i)) can be calculated according to equations (4), (5) and (6) respectively.

[0094]

[0095] Where i = 1, 2, ..., n, h i Represents the path P i The specific usage rules of ETX routing metrics are as follows:

[0096] (a) First, calculate the ETX sum, ETX mean, and mean square error required for each path from node c through each neighbor node to the destination node.

[0097] That is {ETX(i),(i=1,2,…,n)}, and {σ ETX (i),(i=1,2,…,n)}.

[0098] (b) Sort {ETX(i), (i=1,2,…,n)} from small to large, and take the paths corresponding to the first three smallest ETX and values ​​to form a set of candidate paths. All paths with less than three ETX and values ​​are selected into the set of candidate paths.

[0099] (c) In the set of candidate paths, the path corresponding to the minimum ETX mean square error can be selected as the optimal path. Assume σ ETX (f) is the minimum value, then path P f (From node c through neighbor node f to the destination node) is the optimal path, so neighbor node f is the next hop node.

[0100] This method combines the ETX sum (ETX(i)) and the ETX mean and ETX mean square error (σ ETX (i)) The combined use method not only ensures the link quality, but also avoids the situation where there is a large ETX link in the selected optimal path to a certain extent.

[0101] In order to facilitate the use of ETX in conjunction with other routing metrics, σ ETX (i) Perform normalization, as shown in equations (7) and (8).

[0102]

[0103] (4) Delay (D)

[0104] Similar to ETX, according to the traditional delay calculation method, the delay on a path is equal to the sum of the delays required by all links on the path, and the path corresponding to the minimum delay is the optimal path. This optimal path selection method is likely to select a path containing a link with a larger delay as the optimal path. For this reason, the embodiment of the present invention proposes a method of using the mean square error of the delay of each link on the path and the sum of the delays. The path P from node c to the destination node via neighbor node i i The delay sum (D(i)) and the delay mean and the delay mean square error (σ D (i)) can be calculated according to equations (9), (10) and (11) respectively.

[0105]

[0106] Where i = 1, 2, ..., n, h i Represents the path P i The specific usage rules of the delay routing metric are as follows:

[0107] (a) First, calculate the delay and delay mean and mean square error of each path from node c to the destination node through each neighbor node. That is, {D(i), (i=1,2,…,n)}, and {σ D (i),(i=1,2,…,n)}.

[0108] (b) Sort the delay sum values ​​{D(i), (i=1,2,…,n)} from small to large, and select the paths corresponding to the first three smallest delay sum values ​​to form a candidate path set. All paths with less than three delay sum values ​​are selected into the candidate path set.

[0109] (c) In the set of candidate paths, the path corresponding to the minimum mean square error of delay can be selected as the optimal path. Assume σ D (f) is the minimum value, then path P f (From node c through neighbor node f to the destination node) is the optimal path, so node f is selected as the next hop node.

[0110] This method combines the delay sum (D(i)) and the delay mean and the delay mean square error (σ D (i)) The combined use method not only ensures the minimum delay of the path, but also avoids the situation where there are links with large delays in the selected optimal path to a certain extent.

[0111] In order to facilitate the use of delay in conjunction with other routing metrics, σ D (i) Perform normalization, as shown in equations (12) and (13).

[0112]

[0113] (5) Hop count (HC)

[0114] The hop count indicates the number of nodes that the path from the neighbor node to the root node passes through. The use of this routing metric can avoid selecting a neighbor node with a large hop count as the next hop node. The hop count is considered in the mean square error calculation of the delay and ETX routing metrics, so the embodiment of the present invention will no longer evaluate the hop count metric of the node separately.

[0115] The above routing metrics have an important impact on the comprehensive evaluation of neighbor nodes in all aspects. CR-T uses a situation-aware method to evaluate the node residual energy index and cache occupancy, and uses the sum, mean and mean square error methods to evaluate ETX and delay, which can be well applied to the construction of comprehensive objective functions.

[0116] 2. Design of membership function for each routing metric

[0117] CR-T determines the membership function of each routing metric through the assignment method. The membership functions corresponding to the above routing metrics are as follows:

[0118] (1) Membership function of REI

[0119] The membership function of REI can be expressed by the inverse tangent membership function, as shown in Equation (14). The more residual energy a neighbor node has, the greater the possibility that the node will be selected as the next hop node, and vice versa.

[0120]

[0121] (2) Membership function of BOR

[0122] The membership function of BOR can be represented by the Gaussian membership function, as shown in Equation (15). The lower the cache occupancy of the neighbor node, the greater the possibility that the node will be selected as the next hop node, and vice versa.

[0123]

[0124] (3) Membership function of ETX

[0125] The membership function of ETX can be represented by the descending half-normal distribution membership function, as shown in Equation (16). The lower the cache occupancy of the neighbor node, the greater the possibility that the node will be selected as the next hop node, and vice versa.

[0126]

[0127] (4) Membership function of time delay

[0128] The membership function of the delay can be expressed by the Gaussian membership function, as shown in formula (17). The smaller the delay, the greater the possibility that the node is selected as the next hop node, and vice versa.

[0129]

[0130] 3. Design of triangle model fusion operator and situational awareness comprehensive objective function

[0131] 1) Design of triangle module fusion operator

[0132] According to the membership functions of each routing metric, a comprehensive membership function including the membership functions of each routing metric can be constructed based on the triangle mode fusion operator. That is, the membership functions of each routing metric are constructed into a comprehensive membership function in the form of a triangle mode fusion operator, which is convenient for application analysis.

[0133] CR-T uses a four-dimensional triangle modulus fusion operator to construct a comprehensive membership function that includes the membership functions of each routing metric, as shown in formula (18). This comprehensive membership function can comprehensively evaluate neighbor nodes.

[0134]

[0135] In formula (18), i represents the i-th neighbor node, g j(i)(j=1,2,3,4,5) represents the membership function of each routing metric of neighbor node i. According to Definition 3, formula (18) also satisfies the characteristics of the triangle modulus fusion operator:

[0136] (1) Dimension reduction mapping: f:[0,1] 4 →[0,1];

[0137] (2) If but

[0138] like but

[0139] (3) If but

[0140] (4)

[0141] (5) Strengthening: If but

[0142] (6) Harmony: If but

[0143]

[0144] in,

[0145] The triangle modulus fusion operator can increase the probability of excellent nodes being selected as the next hop and reduce the probability of inferior nodes being selected. In addition, the application of the triangle modulus operator can also reconcile the contradictions between the routing metrics of neighboring nodes and balance the next hop node selection criteria.

[0146] 2) Design of comprehensive objective function for situational awareness

[0147] According to the comprehensive membership function formula (18) and the maximum membership principle, the context aware comprehensive objective function (CA-OF) can be designed as shown in formula (19).

[0148]

[0149] The objective function is the basis for obtaining and updating routing metric information, calculating node rank values, building and maintaining network topology, and selecting the optimal path. The context-aware comprehensive objective function designed in the embodiment of the present invention can comprehensively evaluate the REI, BOR, ETX, delay and hop count of neighboring nodes based on the triangle module fusion operator and the maximum membership principle. In this way, the optimal path for data transmission can be obtained to improve network performance.

[0150] 4. Calculation of node rank and design of next-hop node selection mechanism

[0151] 1) Node rank calculation design

[0152] The rank of a node represents the position of the node relative to the root node in the network. To avoid loops, the rank value of the node must be strictly increasing from the root node to the leaf node. The node rank can be obtained based on the objective function. The embodiment of the present invention proposes a new method for calculating the node rank based on the context-aware comprehensive objective function.

[0153] The rank value of the root node is set to 1.0. The ranks of other non-root nodes are calculated according to formula (20), where c is a common non-root node, i is the neighbor of c, and R c (i) indicates the rank value corresponding to node c when node c selects i as the next hop, R cp (i) is the rank value of node i. If c has n neighbor nodes, then {R c (1),R c (2),…,R c (n)}; if min{R c (1),R c (2),…,R c (n)}=R c (f), then node f can be selected as the next hop of c.

[0154] R c (i) = R cp (i)+(OF CA (i)+1),i=1,2,…,n (20)

[0155] 2) Design of next-hop node selection mechanism

[0156] According to formula (20), node c can obtain its own rank value on the path to the destination node through each neighbor node, and then select the node corresponding to the minimum rank value as the next hop. However, if the following special cases exist, special processing is required:

[0157] (1) If the rank calculated by a neighbor node is less than the rank calculated by the current next-hop node, but the difference between the two is less than the next-hop node replacement threshold, the current next-hop node will continue to be used to transmit data without changing the next-hop node. The use of this mechanism can effectively ensure the stability of the network topology without affecting network performance.

[0158] (2) If the calculated rank value is less than or equal to 1 or greater than the total number of nodes in the network, it means that the rank value calculated through the neighbor node is wrong or the quality of the path is extremely poor or the processing capacity of the nodes on the path is extremely low, and it should be recalculated; if the result is still the same, the node is directly set as a leaf node.

[0159] (3) If there are multiple neighbor nodes at the same time, and the rank values ​​of node c obtained through their paths are the smallest and equal, then the node with the largest set of neighbor nodes can be selected as the preferred parent node. This is because the larger the set of neighbor nodes, the larger the selection range of the next-hop node, and the greater the probability of selecting the optimal node as the next-hop node.

[0160] (4) If node c has only one neighbor, node c needs to wait for a while to obtain more neighbor nodes. If the number of neighbor nodes of node c is greater than or equal to 2, node c selects the next hop node by executing the CR-T method. If the number of neighbors of node c is still 1, node c directly selects the neighbor as the next hop without executing the CR-T method to save energy and other resources. At this time, the rank of node c is equal to the rank value of the neighbor plus 1.

[0161] CR-T comprehensively, objectively and efficiently evaluates the possibility of each neighbor node becoming the next hop node based on theories such as the triangle modulus fusion operator and the maximum membership principle, and finally selects the optimal path to transmit data. It solves the problems of the existing RPL and its related improved protocols in that the routing metrics are not comprehensive, the routing metrics cannot perceive the path conditions of the uplink and downlink links, the lack of a scientific theoretical system for multi-routing metric evaluation, and the simple node rank value and next hop node selection mechanism. Therefore, compared with the existing RPL and its related improved methods, CR-T can significantly improve network performance and better meet the requirements of various application scenarios.

[0162] What are the key technical points and pre-protection points of the embodiments of the present invention?

[0163] (1) A context-aware calculation mechanism for uplink and downlink node residual energy index and cache occupancy rate;

[0164] (2) ETX and delay are comprehensively calculated using their sum, mean and mean square error values;

[0165] (3) Design of membership functions of each routing metric and four-dimensional triangular modulus fusion operator;

[0166] (4) Design of comprehensive objective function of situation awareness;

[0167] (5) Calculation of node rank and design of next-hop node selection mechanism.

[0168] Those skilled in the art will appreciate that the accompanying drawing is only a schematic diagram of a preferred embodiment, and the serial numbers of the embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0169] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A new low-power lossy network routing method, characterized in that: Context-aware path conditions and scientifically evaluating multiple routing metrics, the method includes: According to the membership functions of each routing metric, a comprehensive membership function including each routing metric is constructed based on the triangle module fusion operator; Design the situational awareness comprehensive objective function based on the comprehensive membership function and the maximum membership principle; Design a node rank calculation mechanism based on the context-aware comprehensive objective function, and select the optimal path based on the calculation mechanism of the new next-hop node; This routing method is used in low-power lossy network environments.

2. A new low-power lossy network routing method according to claim 1, characterized in that: The comprehensive membership function is: In the formula, i represents the i-th neighbor node, g j (i) represents the membership function of each routing metric of neighbor node i.

3. A new low-power lossy network routing method according to claim 1, characterized in that: The situational awareness comprehensive objective function is: OF CA =min(OF CA (i)) 4. A new low-power lossy network routing method according to claim 1, characterized in that: The calculation mechanism of the node rank is: 1) Node rank calculation design The rank value of the root node is set to 1.0, and the ranks of other non-root nodes are calculated. c is a common non-root node, i is the neighbor of c, and R c (i) indicates the rank value corresponding to node c when node c selects i as the next hop, R cp (i) is the rank value of node i. If c has n neighbor nodes, find {R c (1),R c (2),…,R c (n)}; if min{R c (1),R c (2),…,R c (n)}=R c (f), select node f as the next hop of c; R c (i)=R cp (i)+(OF CA (i)+1),i=1,2,…,n 2) Design of next-hop node selection mechanism (1) If the rank calculated by a neighbor node is smaller than the rank calculated by the current next-hop node, but the difference between the two is smaller than the next-hop node change threshold, the current next-hop node will continue to be used to transmit data without changing the next-hop node; (2) If the calculated rank value is less than or equal to 1 or greater than the total number of nodes in the network, it means that the rank value calculated through the neighbor node is wrong or the path quality is extremely poor or the node processing capacity on the path is extremely low, and it should be recalculated; if the result is still the same, the node is directly set as a leaf node; (3) If there are multiple neighbor nodes at the same time, and the rank value of node c obtained through their paths is the smallest and equal, then the node with the largest set of neighbor nodes is selected as the preferred parent node; (4) If node c has only one neighbor, node c needs to wait for a while to obtain more neighbor nodes. If the number of neighbor nodes of node c is greater than or equal to 2, node c selects the next hop node by executing the CR-T method. If the number of neighbors of node c is still 1, node c directly selects the neighbor as the next hop, and the rank of node c is equal to the rank value of the neighbor plus 1.

Images