A fault-tolerant reconfigurable routing strategy based on a priori knowledge base

Abstract

The present invention relates to the technical field of routing in self-organizing networks, and provides a fault-tolerant reconfigurable routing strategy based on a priori knowledge base, which includes a process of periodically maintaining a location information base for each node in the network, building a priori knowledge base using the location information base, and making an optimal decision on the selection of the next transmission link using the priori knowledge base; it also includes the execution process of the fault-tolerant strategy and the execution process of the reconfigurable strategy. The fault-tolerant reconfigurable routing strategy based on the priori knowledge base of the present invention is task-oriented, based on the node information knowledge base, dynamically schedules related nodes to measure and analyze network resources, and realizes on-demand link reconstruction and route selection; effectively avoids congested nodes or areas during route selection, so that the network load is more evenly distributed to appropriate surrounding nodes, optimizes the message delivery rate and transmission delay, selects the best transmission path in real time, and ensures the rapid switching and self-healing capabilities of the route.

Description

Technical Field

[0001] The invention relates to the technical field of self-organizing network routing, in particular to a fault-tolerant reconfigurable routing strategy based on a priori knowledge base. Background Art

[0002] The network structure without infrastructure is usually called an ad hoc network. Multiple mobile nodes form a multi-hop wireless network. Each node has the function of a router. The network organization is temporary, on-demand, and automatic. Wireless mobile ad hoc networks have the characteristics of multi-hop networks, dynamic changes in network topology, distributed control, peer-to-peer, temporary, self-organizing, limited link bandwidth, unidirectional channels, limited energy, and limited security.

[0003] Based on the characteristics of this network, self-organizing network routing technology not only needs to consider the number of routing hops, but also needs to comprehensively consider technical indicators such as the channel's signal-to-noise ratio, latency, packet loss rate, equipment load and power.

[0004] At present, the research on ad hoc network routing technology mainly includes the following:

[0005] 1) Popular Routing Protocol

[0006] Routing protocols for intermittent networks are called epidemic routing. This protocol relies on the theory of epidemic algorithms. When two nodes meet, they exchange their message lists. After the exchange, each node knows whether the other node has a message that it does not have. If not, the node requests the message from that node. This means that as long as there is space in the buffer, messages will be exchanged between nodes in the network when they meet.

[0007] 2) Routing based on historical messages

[0008] This method is a slope routing method. Nodes with higher levels are more active near the base station and have more opportunities to communicate with the base station. However, there is a "slow start" problem. Especially in a large-scale network environment, relays that have the opportunity to communicate with the source node may be far away from the target node and may not have the opportunity to communicate with the target node. Therefore, it is impossible to have a high success rate to be selected as a relay. In this way, it is difficult for the source node or other relays to make correct judgments when forwarding to the next hop, and intelligent decisions need to be made based on the message forwarding history.

[0009] 3) Controllable mobile routing

[0010] A series of special mobile nodes called information ferries (or short-term ferries) are used to provide communication services in the node scheduling area. The concept of non-randomness of node mobility is introduced and this non-randomness is used to assist in forwarding data. Two MF variations depending on the initial forward motion of the ferry and the node are studied. The design uses mobility to improve data transmission performance and reduce node energy consumption, and its performance superiority is verified in a variety of different network environments.

[0011] 4) Encoding-based routing

[0012] The basic idea is to apply network coding to the forwarded data packets to generate new data packets and corresponding coding vectors. The coding vectors are transmitted together with the data packets. When a sufficient number of data packets are received, the destination node can decode and restore the original data packets. Routing based on network coding can send messages to nodes in the network with a high probability with very little overhead, which is particularly suitable for restricted network environments.

[0013] Although a lot of research, simulation and analysis have been carried out on relevant application scenarios and key technologies at home and abroad, and a certain amount of software algorithms have been accumulated, there are still many problems. Mainly reflected in:

[0014] First, related research has basically remained at the theoretical analysis stage, without considering the actual engineering application constraints, making it difficult to implement engineering in actual networking;

[0015] Secondly, the communication environment of ad hoc networks is relatively harsh, and link disconnection and recovery are sudden and random. The use of long-distance communication or known node mobility will lead to rapid consumption of node limited battery, low data transmission rate and large delay. Summary of the invention

[0016] The purpose of the present invention is to provide a fault-tolerant reconfigurable routing strategy based on a priori knowledge base to solve the problems raised in the above background technology.

[0017] The fault-tolerant reconfigurable routing strategy based on the prior knowledge base includes:

[0018] Step 1: The step 1 includes periodically maintaining a location information base for each node in the network, building a priori knowledge base using the location information base, and making an optimal decision on the selection of the next transmission link using the priori knowledge base;

[0019] Step 2, the step 2 includes the execution process of the tolerance strategy; the tolerance strategy includes counting the message sending status of each node within a specified time period, obtaining the distribution status of the network load, and selecting the node with a large idle ratio of the egress message queue as the next hop node of the route according to the distribution status of the network load;

[0020] Step three, the step three includes the execution process of the reconfigurable strategy; the reconfigurable strategy includes routing reconfiguration based on quality of service (QoS), and each node calculates the optimal path for the next hop of the data message in real time based on the quality of service (QoS) data transmission requirements.

[0021] Furthermore, in the step one, the construction process of the prior knowledge base includes: setting up a master node and slave nodes in the network, and the location information library records the geographic location information, running speed, and running trajectory of any node in the network; the slave node periodically reports its own geographic location information, and the master node not only periodically sends its own location information to all slave nodes, but also shares the other slave node information it collects with all slave nodes in the network.

[0022] Furthermore, the specific process of step one includes:

[0023] Define the various parameters of the location information library:

[0024] Define Node O i →O j The link transmission delay between i→j , node O i →O j The multi-hop path between i→j , from which we can calculate node O i →O j The total transmission delay of the link is

[0025] Where Q i , P i The data packets are at node O. i The queuing delay and data processing delay of forwarding;

[0026] For a selected node O i , Q i , P i Considered as a constant value C i , so the above formula can be further simplified to

[0027] Define Node O i →O j The link connection time between them is:

[0028]

[0029] Where a = v i cosθ i -v j cosθ j ; b = x i -x j ;

[0030] c = v i sinθ i -v j sinθ j ; d = y i -y j ; v i and v j are the average velocities of nodes O i and O j respectively; θ i and θ j are the moving directions; (x i , y i ), (x j , y j ) are the position coordinates corresponding to nodes O i and O j respectively;

[0031] Define the path cost parameter between nodes O i →O j as where ζ, ε are adjustment coefficients used to adjust the relative importance of the two time parameters in the link; in the selection of the next transmission link, use the information in the prior knowledge base to select the path with a smaller path cost for optimal routing selection.

[0032] Furthermore, the specific process of step two includes:

[0033] S1: Set up an outgoing message queue for each node in the network, and all messages of each node are processed and sent through this message queue;

[0034] S2: Initialize the same length of the outgoing message queue, and use a global variable counter to count the length occupied by the current queue;

[0035] S2: Whenever a message is added to the queue, the counter is incremented by one, and conversely, when a message is dequeued and processed, the counter is decremented by one;

[0036] S3: Define the idle ratio R of the outgoing message queue of the node:

[0037] In the formula: U represents the occupied length of the queue; L is the total length of the queue;

[0038] Set the congestion threshold to t to specify the minimum length of the idle queue, and its value is configured according to the network carrying capacity; if R < t, it means there is data backlog in this link, and then transfer to S4;

[0039] S4: Select and execute one of case A, case B, and case C;

[0040] Case A: If the link is not interrupted, the following steps should be followed: the node puts the data to be sent into a temporary storage. When the amount of temporarily stored data reaches a certain threshold, the node's sending and receiving priority is increased, the geographic location reporting process is started, and the phased array antenna beam pointing is updated, so that data can continue to be sent. In this case, there is no need to find a new routing path again;

[0041] Case B: If the optimal path link is interrupted, this is a link failure. The node deletes the existing optimal routing path and sends data according to the suboptimal path in the routing table. At the same time, it initiates the routing recovery process to find a new optimal path.

[0042] Case C: If both the optimal path and the suboptimal path are disconnected, this is a link failure. In this case, you need to dynamically select a link with the largest idle ratio of the egress queue for forwarding.

[0043] Further, in S4, the specific selection process is as follows:

[0044] S41: Check the topological status of the current node and all adjacent nodes;

[0045] S42: Calculate the idle ratio R of the egress queue from the current node to all adjacent nodes i , determine the largest one R next =max{R i};

[0046] S43: If the link is not interrupted, proceed with the process as in case A, otherwise continue searching;

[0047] S44: If all links are disconnected, the node stores the data in a permanent memory and waits to establish a new link before sending it again.

[0048] Furthermore, the specific process of step three includes:

[0049] P1: Constructing a network QoS routing model: Given a network G =<V、E、P> ,V is the node set, E is the edge (link between nodes) set, and P is the path set; the basis for selecting the optimal path is to select a path p∈P from s to t for the source node s∈V and the destination node t∈V, so that the optimal QoS index is met;

[0050] P2: QoS factor selection: select channel quality parameter cqi, hop number parameter hop, auxiliary, where:

[0051] P21: The signal quality parameter cqi processing flow is as follows:

[0052] If cqi<CQI_MIN, then θ cqi =0;

[0053] If CQI_MIN≤cqi<CQI_MAX, then

[0054] If cqi ≥ CQI_MAX, then θ cqi =1;

[0055] P22: The hop count parameter hop processing flow is as follows:

[0056] If hop<Hop_Threshold, then

[0057] Where Hop_Threshold>1;

[0058] If hop ≥ Hop_Threshold, then θ hop =0;

[0059] P23: The selection process of auxiliary QoS factors is as follows:

[0060] P231: Develop auxiliary parameters based on node load, link capacity, and latency, and determine the proportion of parameter factors based on the importance of different parameters;

[0061] P232: Based on the idea of ​​linear normalization, by assigning different weights to different auxiliary QoS factors, a normalized conversion function is used to reduce multiple auxiliary QoS factors to the path weights for calculating routing; the conversion function is as follows:

[0062]

[0063] in, To pre-set the proportional weight of the auxiliary QoS metric factor, QoS i For each auxiliary QoS factor value, the above formula satisfies:

[0064]

[0065] P24: Make path selection decisions based on QoS metrics; specifically: the larger the QoS value, the better the corresponding path, and the higher the priority of the corresponding path selection;

[0066] The calculation formula of QoS metrics is as follows:

[0067] QoS = α × θ cqi +β×θ hop +δ×θ other .

[0068] Compared with the prior art, the present invention has the following beneficial effects:

[0069] First, the present invention provides a reliable routing selection and fast dynamic switching method in a self-organizing network environment. The method is based on a fault-tolerant reconfigurable routing strategy based on a priori knowledge base, is task-oriented, and is based on a node information knowledge base. It dynamically schedules related nodes to perform network resource measurement and analysis to achieve on-demand link reconstruction and routing selection.

[0070] Second, the prior knowledge base strategy targets the predictability of the movement trajectory of nodes and topological connection changes in ad hoc networks. It can use the prior knowledge of the network to select the message forwarding path, improve the performance of the routing protocol, and thus enhance the adaptability to changes in network status.

[0071] Third, through the tolerance strategy based on the idle ratio of the node export message queue, the congestion status of each node in the network can be effectively estimated, so as to effectively avoid congested nodes or areas during route selection, so that the network load is more evenly distributed to appropriate surrounding nodes, and the message delivery rate and transmission delay are optimized.

[0072] Fourth, provide a reconfigurable routing strategy based on QoS to dynamically adjust the routing according to the available network resources. Quality of service (QoS) is a series of service requests that the network must meet when transmitting data streams. It can be quantified into performance indicators such as bandwidth, latency, latency jitter, loss rate, throughput, and cost, so as to select the best transmission path in real time and ensure the rapid switching and self-healing capabilities of the routing. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 The flowchart of the fault-tolerant reconfigurable routing strategy based on a priori knowledge base of the present invention. DETAILED DESCRIPTION

[0074] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0075] See also Figure 1, a disconnection-tolerant reconfigurable routing strategy based on a priori knowledge base, comprising: step one, the step one comprising periodically maintaining a location information base for each node in the network, and using the location information base to construct a priori knowledge base, and using the priori knowledge base to make the optimal decision on the selection of the next transmission link; step two, the step two comprising the execution process of the disconnection-tolerant strategy; the disconnection-tolerant strategy comprises counting the message sending conditions of each node within a specified time period, obtaining the distribution of the network load, and selecting a node with a large idle ratio of the export message queue as the next hop node of the routing according to the distribution of the network load; step three, the step three comprising the execution process of the reconfigurable strategy; the reconfigurable strategy comprises reconfiguring the routing according to the quality of service (QoS), and each node calculates the optimal path for the next hop of the data message in real time according to the quality of service (QoS) data transmission requirements.

[0076] In step one, the prior knowledge base strategy is adopted.

[0077] During operation, each node in the network periodically maintains a location information database, which mainly records the geographical location information, running speed, etc. of any node in the network, and calculates the running trajectory of the node. All slave nodes periodically report their own geographical location information. In addition to periodically sending its own location information to all slave nodes, the master node also notifies all slave nodes in the network of the information it collects from other slave nodes.

[0078] Use the location information library to define the various parameters of the location information. Define node O i →O j The link transmission delay between i→j , node O i →O j The multi-hop path between i→j , from which we can calculate node O i →O j The total transmission delay of the link is

[0079] Where Q i , P i The data packets are at node O. i The queuing delay and data processing delay of forwarding.

[0080] For a selected node O i , Q i , P i Considered as a constant value C i , so the above formula can be further simplified to

[0081] Define Node O i →O j The link connection time between them is:

[0082]

[0083] Where a = v i cosθ i -v j cosθ j ; b = x i -x j ;

[0084] c=v i sinθ i -v j sinθ j ; d = y i -y j ;v i and v j Node O i and O j The average speed of i and θ j is the moving direction; (x i ,y i )、(x j ,y j ) are node O i and O j The corresponding position coordinates.

[0085] Define Node O i →O j The path cost parameter between Where ζ and ε are adjustment coefficients, which can adjust the relative importance of the two time parameters in the link according to the situation.

[0086] In the selection of the next transmission link, the prior knowledge of the location information database is used to select a path with less path cost for optimized routing selection, which can effectively reduce the queuing delay and data processing delay of data forwarding, while ensuring the high efficiency of the connection between nodes.

[0087] In step 2, a tolerant failure strategy is adopted.

[0088] The outage tolerance strategy obtains the distribution of network load by counting the message sending status of each node within a period of time, and selects the node with a large idle ratio of the export message queue as the next hop node of the route. This is because the node with a larger idle ratio is less likely to consume resources and the possibility of sending congestion or interruption is lower.

[0089] Each node in the network maintains an outgoing message queue, and all messages are processed and sent through this message queue. The same length of the outgoing message queue is initialized, and a global variable counter is used to count the length occupied by the current queue. Whenever a message is put into the queue, the counter is incremented, and conversely, when a message is taken out of the queue for processing, the counter is decremented. For a link with normal message sending and receiving, the queue length will continuously return to 0; when there is link congestion or link interruption, the queue length will continuously accumulate until it exceeds a certain threshold.

[0090] Define the idle ratio R of the outgoing message queue of the node:

[0091] In the formula: U represents the occupied length of the queue; L is the total length of the queue.

[0092] Define the congestion threshold as t to specify the minimum length of the idle queue, and its value is configured according to the network carrying traffic. If R < t, it means there is data backlog on this link. At this time, it needs to be processed in the following three cases.

[0093] Case A: The link is not interrupted;

[0094] This case conforms to the characteristics of the fault-tolerant link and should be processed according to the following steps: The node puts the data to be sent into the temporary memory. When the amount of temporarily stored data reaches a certain threshold, the sending and receiving priority of the node is increased, the geographical location reporting process is started, and the phased array antenna beam pointing is updated, so that the data can continue to be sent. In this case, there is no need to find a new routing path.

[0095] Case B: The optimal path link is interrupted;

[0096] This case belongs to a link failure, then the node deletes the existing optimal routing path, sends data according to the sub-optimal path in the routing table, and at the same time initiates a routing recovery process to find a new optimal path.

[0097] Case C: Both the optimal path and the sub-optimal path are interrupted;

[0098] This case belongs to a link failure. At this time, it is necessary to dynamically select a link with the largest idle ratio of the outgoing queue in one hop for forwarding.

[0099] The specific selection process is as follows:

[0100] Check the topological status of the current node and all adjacent nodes;

[0101] Calculate the idle ratio R of the outgoing queue from the current node to all adjacent nodes i , and determine the largest one R next = max{R i};

[0102] If the link is not interrupted, proceed as per situation A; otherwise, continue searching.

[0103] If all links are disconnected, the node stores the data in permanent storage and waits to establish a new link before sending it again.

[0104] Step 3: Adopt a reconfigurable strategy;

[0105] The present invention performs routing reconfiguration according to quality of service (QoS), and each node calculates the optimal path for the next hop of the data message in real time according to the demand of QoS data transmission.

[0106] The network QoS routing model is constructed as follows: Given a network G =<V、E、P> ,V is the node set, E is the edge (link between nodes) set, and P is the path set. The basis for selecting the optimal path is to select a path p∈P from s to t for the source node s∈V and the destination node t∈V so that the optimal QoS index is met.

[0107] When the present invention performs QoS factor selection, a method of combining the selection of important QoS factors and auxiliary QoS factors is adopted.

[0108] For the selection of important QoS factors, the channel signal quality and the number of path hops are considered as important QoS factors. For communication links, the lower the channel signal quality, the lower the service transmission rate; the more hops, the longer the delay, and communication cannot proceed normally. Therefore, when meeting business needs, ensuring channel quality while shortening the number of path hops can expand network throughput, improve network performance and overall efficiency. Therefore, when selecting QoS parameters, focus on selecting the channel quality parameter cqi and the hop number parameter hop.

[0109] The signal quality parameter cqi processing flow is as follows:

[0110] If cqi<CQI_MIN, then θ cqi =0;

[0111] If CQI_MIN≤cqi<CQI_MAX, then

[0112] If cqi ≥ CQI_MAX, then θ cqi =1.

[0113] The hop count parameter hop processing flow is as follows:

[0114] If hop<Hop_Threshold, then

[0115]

[0116] Where Hop_Threshold>1;

[0117] If hop ≥ Hop_Threshold, then θ hop =0.

[0118] For the selection of auxiliary QoS factors, auxiliary parameters such as node load, link capacity, delay, etc. can be formulated according to actual applications, and the proportion of parameter factors can be adjusted according to the importance of different parameters. Based on the idea of ​​linear normalization, different auxiliary QoS factors are assigned different weights, and a normalized conversion function is used to reduce multiple auxiliary QoS factors to the path weights for calculating routing. The conversion function is as follows:

[0119]

[0120] in, To pre-set the proportional weight of the auxiliary QoS metric factor, QoS i For each auxiliary QoS factor value, the above formula satisfies:

[0121]

[0122] The larger the QoS value, the better the corresponding path;

[0123] The QoS metrics of the present invention are as follows:

[0124] qos=α×θ cqi +β×θ hop +δ×θ other .

[0125] In addition, the frequency of route updates and the size of messages can be adjusted adaptively to make a reasonable compromise between route overhead and accuracy. By determining whether the change in the state information value exceeds a predetermined threshold, route updates are performed only when the time exceeds the threshold, and existing routes are maintained as much as possible, thereby reducing computational overhead and performance jitter.

[0126] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A fault-tolerant reconfigurable routing method based on a priori knowledge base, characterized in that: It includes: Step 1, which includes the process of periodically maintaining the location information database for each node in the network, constructing a prior knowledge base using the location information database, and making an optimal decision on the selection of the next transmission link using the prior knowledge base; Step 2, which includes the execution process of the fault tolerance strategy; the fault tolerance strategy includes counting the message sending situations of each node within a specified time period, obtaining the distribution status of the network load, and selecting the node with a large idle ratio of the outgoing message queue as the next-hop routing node according to the distribution status of the network load; Step 3, which includes the execution process of the reconfiguration strategy; the reconfiguration strategy includes performing routing reconfiguration according to the quality of service (QoS), and each node calculates the optimal path for the next hop of the data packet in real time according to the QoS data transmission requirements; In the above Step 1, the construction process of the prior knowledge base includes: setting a master node and slave nodes in the network, and the location information database records the geographical location information, running speed, and running trajectory of any node in the network; the slave nodes periodically report their own geographical location information, and in addition to periodically sending its own location information to all slave nodes, the master node also shares the information of other slave nodes it collects to all slave nodes in the network; The specific process of the above Step 1 includes: Defining each parameter of the location information database: Define Node O i →O j The link transmission delay between i→j , node O i →O j The multi-hop path between i→j , from which we can calculate node O i →O j The total transmission delay of the link is Where Q i , P i The data packets are at node O. i The queuing delay and data processing delay of forwarding; For a selected node O i , Q i , P i Considered as a constant value C i , so the above formula can be further simplified to Define Node O i →O j The link connection time between them is: where a = v i cosθ i -v j cosθ j ; b = x i -x j ; c=v i sinθ i -v j sinθ j ; d = y i -y j ;v i and v j Node O i and O j The average speed of i and θ j is the moving direction; (x i ,y i )、(x j ,y j ) are node O i and O j The corresponding position coordinates; Define Node O i →O j The path cost parameter between Among them, ζ and ε are adjustment coefficients, which are used to adjust the relative importance of the two time parameters in the link. In the selection of the next transmission link, the information of the prior knowledge base is used to select the path with the smallest path cost for optimized routing selection. The specific process of the above Step 2 includes: S1: Set an outgoing message queue for each node in the network, and all messages of each node are processed and sent through this message queue; S2: Initialize the same length of the outgoing message queue, and use a global variable counter to count the occupied length of the current queue; S2: Whenever a message is added to the queue, the counter is incremented by one, and conversely, when a message leaves the queue and is processed, the counter is decremented by one; S3: Define the node's egress message queue idle ratio R: In the formula: U represents the occupied length of the queue; L is the total length of the queue; Set the congestion threshold as t to specify the minimum length of the idle queue, and its value is configured according to the network carrying capacity; if R < t, it means that there is data backlog on this link, and at this time, it transfers to S4; S4: Select and execute one of Case A, Case B, and Case C; Case A: If the link is not interrupted, it should be processed according to the following steps: The node puts the data to be sent into the temporary memory. When the amount of temporarily stored data reaches a certain threshold, the transceiver priority of the node is increased, and the geographical location reporting process is started to update the phased array antenna beam pointing, so that the data can continue to be sent; In this case, there is no need to find a new routing path; Case B: If the optimal path link is interrupted, this situation belongs to a link failure, then the node deletes the existing optimal routing path, sends data according to the sub-optimal path in the routing table, and at the same time initiates a routing recovery process to find a new optimal path; Case C: If both the optimal path and the sub-optimal path are interrupted, this situation belongs to a link failure, and at this time, it is necessary to dynamically select a link with the largest idle ratio of the outgoing queue for one hop for forwarding; In the above S4, the specific selection process is as follows: S41: Check the topological status of the current node and all adjacent nodes; S42: Calculate the idle ratio R of the egress queue from the current node to all adjacent nodes i , determine the largest one R next =max{R i }; S43: If this link is not interrupted, it is processed with reference to Case A, otherwise continue to search; S44: If all links are disconnected, the node stores the data in a permanent memory and waits to send it again after a new link is established; The specific process of step three includes: P1: Constructing a network QoS routing model: Given a network G =<V、E、P> ,V is the node set, E is the edge set, that is, the link set between nodes; P is the path set; the basis for selecting the optimal path is to select a path p∈P from s to t for the source node s∈V and the destination node t∈V, so that the optimal QoS index is met; P2: QoS factor selection: select channel quality parameter cqi, hop number parameter hop, auxiliary, where: P21: The signal quality parameter cqi processing flow is as follows: If cqi<CQI_MIN, then θ cqi =0; If CQI_MIN≤cqi<CQI_MAX, then If cqi ≥ CQI_MAX, then θ cqi =1; P22: The hop count parameter hop processing flow is as follows: If hop<Hop_Threshold, then Where Hop_Threshold>1; If hop ≥ Hop_Threshold, then θ hop =0; P23: The selection process of auxiliary QoS factors is as follows: P231: Develop auxiliary parameters based on node load, link capacity, and latency, and determine the proportion of parameter factors based on the importance of different parameters; P232: Based on the idea of ​​linear normalization, by assigning different weights to different auxiliary QoS factors, a normalized conversion function is used to reduce multiple auxiliary QoS factors to the path weights for calculating routing; the conversion function is as follows: in, To pre-set the proportional weight of the auxiliary QoS metric factor, QoS i For each auxiliary QoS factor value, the above formula satisfies: P24: Make path selection decisions based on QoS metrics; specifically: the larger the QoS value, the better the corresponding path, and the higher the priority of the corresponding path selection; The calculation formula of QoS metrics is as follows: QoS=α×θ cqi +β×θ hop +δ×θ other 。

Images