KR101505627B1 - Method for generating subgraph based on quality of service and system thereof - Google Patents

Abstract

The present invention relates to a method for generating a subgraph on quality of service (QoS) and to a system thereof. According to an embodiment of the present invention, the present invention includes the steps of collecting statistical data including QoS from nodes of a network; calculating a service quality metric for generating a sub graph by using the collected statistic data; calculating a selection probability from a previously selected first node in consideration of energy costs for nodes included in the service quality metric; selecting a second node which is next to the first node according to the calculated selection probability; calculating the selection probability, and generating the sub graph by performing the calculation of the selection probability and the selection of the second node from a source node to a target node.

Description

[0001] DESCRIPTION [0002] SERVICE QUALITY BASED AUXILIARY GRAPH CREATION METHOD AND SYSTEM [0003]

More particularly, the present invention relates to an auxiliary graph generation method based on quality of service (QoS) based on quality of service (QoS) and capable of reducing energy costs of the entire network and improving network efficiency. Lt; / RTI >

The present invention has been derived from research carried out as a part of the project for supporting the IT convergence advanced manpower course of the Future Creation Department and the IT Industry Promotion Agency [assignment number: H0401 -13-1004 , title: M2M-based intelligent autonomous production machine research ].

A wireless mesh network is defined as a network in which a large number of nodes are placed in a target area and operate organically. In the wireless mesh network, each node collects, processes, and transmits information, and the nodes located in the middle serve as a router for retransmitting the received message.

In recent years, researches have been conducted to collect and process various information generated in the field by applying a wireless mesh network to industrial sites.

The wireless network applied to the industrial field is called an industrial wireless network.

In the industrial wireless network, there are source routing algorithms and graph routing algorithms for the message transmission path of each node.

The source routing algorithm is a method for the sender to determine the path to the destination. The graph routing algorithm is an algorithm for defining the nodes as vertices and edges and searching for the shortest path from the source to the destination .

Graph Routing is a routing protocol that runs on network-based ISA100.11a. The graph routing algorithm guarantees security, fault tolerance, and scalability, but unlike source routing, there is no full path. Therefore, in the case of vertices (nodes) having many connection points among the vertices (nodes), there is a problem that the energy consumption and the real-time property can not be satisfied due to redundant paths.

The International Electrotechnical Commission (IEC) has recently approved ISA100.11a for wireless system-process control and related applications for industrial automation.

ISA100.11a is an application of wireless sensor networks that can provide reliable and secure operation for management control, notification, and non-critical monitoring in the industry.

ISA100.11a is a protocol suite for low data rate wireless connectivity with fixed, portable and mobile devices requiring limited power consumption, system management, gateways and security It defines gateway and security specifications.

ISA100.11a supports monitoring and process control with a wait time of about 100 ms or allowing selective operation of a shorter waiting time and is capable of controlling all OSI layers: the physical layer, the data link layer, the network layer, the transport layer, As shown in Fig.

Algorithms for generating conventional graph routing have considered metrics and characteristics in the adhoc network or the Internet domain, but have not considered quality of service (QoS).

Therefore, there is a need for a method for generating an auxiliary graph that can perform routing considering service quality.

Korean Registered Patent No. 10-1204518 (Registered on November 19, 2011)

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems of the related art, and it is an object of the present invention to provide an auxiliary graph generation method based on a service quality, which can reduce the energy cost of the entire network and improve the network efficiency, System.

Specifically, the present invention generates an auxiliary graph from a source node to a destination node in consideration of quality of service and energy cost at each node constituting the network, thereby reducing the energy cost of the entire network and improving network efficiency.

According to another aspect of the present invention, there is provided a method of generating an auxiliary graph based on quality of service (QoS) based on quality of service from a source node to a destination node, Collecting statistical data including the statistical data; Calculating a service quality metric for generating an auxiliary graph using the collected statistical data; Calculating a selection probability from each of the nodes selected at the previously selected first node in consideration of the energy cost for the nodes included in the quality of service metric; Selecting a second node as a next node in the first node according to the calculated selection probability; And generating the auxiliary graph by performing the step of calculating the selection probability and the step of selecting the second node from the source node to the destination node.

The statistical data may include a list of costs for paths between the source node and the destination node, a list of time delays for the paths, and a list of slots by a sum of links between the two nodes.

The statistical data may include a packet transmission expected energy cost, a packet transmission expected time, a packet transmission expected reliability, and a residual energy of a corresponding node in each of the nodes.

The step of calculating the selection probability may calculate the selection probability from the first node to each of the nodes by further considering the weight for the cost factor, the weight for the time delay factor, and the weight for the slot factor.

Wherein the step of calculating the selection probability comprises: placing a weight for the cost factor, a weight for the time delay factor and a weight for the slot factor in candidate nodes for the first one of the nodes, The selection probabilities from the first node to each of the candidate nodes may be calculated in consideration of the weights and the energy cost.

Wherein the step of calculating the selection probability includes the step of calculating a weight for the cost factor, a weight for the time delay factor, and a weight for the slot factor, so that a particular one of the candidate nodes can be selected as the second node. A determined bias margin may be given, and the weights to which the bias margin is given may be arranged at the specific node.

Further, the method according to the present invention includes the steps of: updating the statistical data using a Q-learning technique after a predetermined period from the time when the auxiliary graph is generated, and regenerating the auxiliary graph using the updated statistical data As shown in FIG.

The auxiliary graph generation system based on quality of service according to an embodiment of the present invention is a supplementary graph generation system from a source node to a destination node. The auxiliary graph generation system includes statistical data collecting statistical data including quality of service (QoS) A data collecting unit; A QoS metric calculation unit for calculating a service quality metric for generating an auxiliary graph using the collected statistical data; A selection probability calculation unit for calculating a selection probability from each of the nodes selected from the previously selected first node in consideration of an energy cost for the nodes included in the service quality metric; A node selector for selecting a second node as a next node in the first node according to the calculated selection probability; And an auxiliary graph generation unit for performing the selection probability calculation by the selection probability calculation unit and the selection of the second node by the node selection unit from the source node to the destination node to generate the auxiliary graph.

According to the present invention, it is possible to reduce the energy cost of the entire network and improve the network efficiency by generating the auxiliary graph from the source node to the destination node in consideration of the quality of service and the energy cost at each node constituting the network.

Also, according to the present invention, the auxiliary graph can be updated by updating the statistical data using a Q-learning approach after a certain period of time after the generation of the auxiliary graph and regenerating the auxiliary graph using the Q- This can improve the reliability of the auxiliary graph.

Further, according to the present invention, in calculating the selection probability for determining the nodes constituting the subsidiary graph, the determination of the specific node is biased by forcing a certain node to apply a bias margin to a given weight , Which can improve security from external attacks.

FIG. 1 is a flowchart illustrating a method of generating an auxiliary graph based on a service quality according to an exemplary embodiment of the present invention. Referring to FIG.
FIG. 2 is a diagram illustrating an example of a connection assistance graph algorithm. FIG.
3 illustrates an example of node selection between a source node and a destination node to illustrate a method according to the present invention.
Figure 4 shows an example of actions and states between a controller and a system.
FIG. 5 shows an example of an auxiliary graph generation time according to the number of nodes to which the present invention is applied.
FIG. 6 shows an example of consumption of memory resources different from the number of nodes to which the present invention is applied.
FIG. 7 shows an example of generation time according to an auxiliary graph size to which the present invention is applied.
FIG. 8 shows an example of the end-to-end delay time according to the number of event samples to which the present invention is applied.
FIG. 9 shows an example of energy analysis according to the number of subsidiary graphs to which the present invention is applied.
FIG. 10 shows a configuration of a service quality-based auxiliary graph generation system according to an embodiment of the present invention.

Other objects and features of the present invention will become apparent from the following description of embodiments with reference to the accompanying drawings.

Preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear.

However, the present invention is not limited to or limited by the embodiments. Like reference symbols in the drawings denote like elements.

Hereinafter, a service quality-based auxiliary graph generation method and system according to an embodiment of the present invention will be described in detail with reference to FIGS. 1 to 10.

FIG. 1 is a flowchart illustrating a method of generating an auxiliary graph based on a service quality according to an exemplary embodiment of the present invention. Referring to FIG.

Referring to FIG. 1, a method according to the present invention collects statistical data including quality of service (QoS) or quality of service from nodes constituting the network (S110).

At this time, the statistical data related to the quality of service may include a list of costs for the paths between the source and destination nodes, a list of time delays for the paths, and a list of slots by the sum of the links between the two nodes The expected energy cost of the packet transmission, the expected packet transmission time, the expected packet transmission reliability, and the residual energy of the corresponding node at each of the nodes.

When the statistical data is collected, the service quality (QoS) metric for generating the auxiliary graph is calculated using the collected statistical data (S120).

The selection probability to each of the nodes in the previously selected node (hereinafter referred to as "first node ") is calculated in consideration of the energy cost for the nodes among the information included in the service quality metric calculated in step S120 ).

In this case, the selection probability may be a selection probability to each of the nodes selectable from the first node to the next node.

Step S130 may calculate the selection probability from the first node to each of the nodes by further considering the weight for the cost factor, the weight for the time delay factor, and the weight for the slot factor.

At this time, in step S130, a weight for the cost factor, a weight for the time delay factor, and a weight for the slot factor are allocated to candidate nodes for the first one of the nodes, and the weighted values and the energy cost are considered The energy cost used in the selection probability may be calculated by calculating the energy cost between the candidate nodes of either the first node or the candidate nodes and the energy cost between the first node and the candidate nodes It may be a percentage of total energy cost.

Further, step S130 may be a method for attacking a malicious node, and may force a bias margin on weights assigned to a specific node to bias a determination of a specific node. That is, in step S130, a predetermined bias margin is applied to each of the weights for the cost factors, the weights for the time delay factors, and the weights for the slot factors so that a particular one of the candidate nodes for the first node can be selected as the next node And assigning the weights to which the bias margin is given to a specific node, the selection probability for a specific node can be calculated.

When the selection probability for each of the nodes is calculated, a next node (hereinafter referred to as a "second node") to generate an auxiliary graph in the first node is selected according to the selection probability for each of the calculated nodes (S140) .

If it is determined that steps S130 and S140 have been performed to the destination node, a supplementary graph between the source node and the destination node is generated using the selected nodes between the source node and the destination node (steps S150 and S160).

Although not shown in FIG. 1, it is possible to perform a process of generating a supplementary graph and regenerating the supplementary graph after a predetermined period of time, and updating the statistical data after a predetermined period to regenerate the supplementary graph .

The auxiliary graph generation method according to the present invention will now be described in detail.

There are various network performance level factors to consider in generating a subgraph. In the present invention, by using concepts from graph theory and connection subgraph, which are often used in data mining, To solve the problem of finding the auxiliary graph in the graph, it is possible to find the optimal connection auxiliary graph by collecting statistics of the entire graph.

Connection assisted graph mining has been introduced in finding a connected auxiliary graph that captures the relationship between two non-neighboring nodes in the graph, for example, the two vertices s, t shown in FIG. 2, Is to find an auxiliary graph that contains s, t and at least one other vertex that can maximize a certain goodness function.

This is done by determining the appropriate function to measure fitness and designing (or inventing) a connection subgraph algorithm.

Before describing the connection subgraph algorithm, the following parameters are defined.

For a given graph, defining a graph representing the interference in the graph as a conflict graph F _G , if two vertices in F _G collide due to interference, then there is an edge between the two vertices.

The traffic load on each link f connecting the vertices in the collision graph F _G can be expressed as Equation (1) below.

[Equation 1]

(F) denotes a capacity of a link, T denotes a number of slots in a period for supporting traffic load, c denotes a capacity of a link, Term.

For graph G, the connected solution subgraph is represented by H as shown in FIG. 2 and the fit function is represented by g (H). A goodness function may be a metric that represents an end to end delay if a minimum delay is required. The fit function may be replaced with throughput. For example, by transmitting a large amount of data in the minimum number of slots of the superframe, that is, w (f), the throughput is maximized.

Therefore, you can find a fit function that minimizes the end-to-end delay and maximizes the throughput. For example, in order to select two fitting functions g (H ₁ ) and g (H ₂ ) between the source node and the destination node s and t, which are the two vertices shown in FIG. 2, ) To w (7) can be used.

The connection auxiliary graph algorithm collects the statistics of the entire graph and outputs the result data thereof.

In this case, the input data of the connection assistance graph algorithm includes a network graph, for example, an edge weighted undirected graph, including information on nodes, link information between nodes, , And the output data of the connection assistance graph algorithm may include a list of costs (L _C ) for paths, a list of time delays for paths (L _T ), and a list of slots by the sum of links between the two nodes (M), and the output data may be data between the source node and the destination node, or may be data using all the nodes.

Input data and output data in the connection assistance graph algorithm (Algorithm 1) can be expressed as shown in Table 1 below.

[Table 1]

Here, W denotes a goodness function for a path, a list of costs (L _C ) for paths, and a list of time delays (L _T ) for paths are arranged in ascending order List, and the list of slots (M) by summing the links between two nodes means a list of accumulated slots that maximizes the fit function.

The auxiliary graph generation method according to the present invention may be an auxiliary graph generation algorithm based on a service quality or an energy, and may generate an auxiliary graph using a connection assist graph algorithm in a network.

This auxiliary graph generation algorithm collects statistics from the system manager, probabilistically evaluates network conditions, and then selects at least one or more nodes to construct the auxiliary graph.

This auxiliary graph generation algorithm (algorithm 2) will be described as follows.

The auxiliary graph generation algorithm calculates statistics from all the nodes in the network, such as the expected energy cost (E _i ), the estimated packet transmission time (T _i ), the expected packet transmission reliability (R _i ) RE _i ).

A metric for generating a probabilistic auxiliary graph using the collected statistics, for example, a service quality metric is calculated as shown in Equation (2) to Equation (4) below.

&Quot; (2) "

&Quot; (3) "

&Quot; (4) "

The parameters and parameters described in the equations (2) to (4) are as shown in the following [Table 2].

[Table 2]

Then, in order to bias the probability of node selection for generating the auxiliary graph, the result data output by the connection assistance graph algorithm, i.e. a list of costs (L _C ) for the paths, a list of time delays for the paths (L _T ), and a list of slots (M) based on the sum of links between the two nodes, and for each of the determined candidate nodes a cost factor weight (w _c ), a time delay factor weight (w _T ), and a slot factor weight (w _w ).

And the weights are candidate nodes, where candidate nodes are nodes that can be selected at the previously selected first node or at the source node, and calculate the selection probability for the candidate nodes.

Here, the selection probability for each of the candidate nodes can be expressed as Equation (5) below.

&Quot; (5) "

That is, the selection probability of each of the candidate nodes selectable at the first node is calculated by using the ratio of the estimated energy cost between the candidate node and the expected energy cost between the candidate node at the first node and the weights of the candidate node at the first node Can be calculated.

The second node, which is the next node, is selected by using the selection probability value for each of the calculated candidate nodes, and this process is repeated from the source node to the destination node.

For example, in order to generate the auxiliary graph from the source node i to the destination node n, as shown in Fig. 3, first of all the selection probabilities P _{i , (k , m)} for generating a subsidiary graph between the source node and the destination node by selecting the _k- node according to the selection probability P _{k , m} among the candidate nodes at the k-th node _{k ,} Node).

The source node, the destination node, and the selected node between the source node and the destination node are used to generate the auxiliary graph between the source node and the destination node.

The generated auxiliary graph may be added to the auxiliary graph list S.

Through this process, it is possible to generate a supplementary graph between the source node and the destination node according to the present invention. That is, the auxiliary graph generated by the present invention includes a list of costs (L _C ) for paths, a list of time delays (L _T ) for paths, a list of slots by sum of links between two nodes ), Considering the packet transmission expected energy cost (E _i ), the packet transmission expected time (T _i ), the packet transmission expected reliability (R _i ) and the residual energy of the node (RE _i ) Energy costs to the nodes can be reduced, thereby reducing the energy costs of the entire network and improving network efficiency. However, in Fig. 3, the energy cost is expressed as Ci, j in order to emphasize that energy is regarded as cost in generating the auxiliary graph. The energy cost Ci, j shown in Fig. 3 is the same concept as Ei, j obtained from the equations (2) to (4).

The auxiliary graph generation method according to the present invention can update the statistical data using a Q-learning approach after a certain period of time after the generation of the auxiliary graph, and regenerate the auxiliary graph using the Q-learning approach. This auxiliary graph regeneration process can improve the reliability of the auxiliary graph.

The goal of this Q-learning is to control the system optimally and can be used to design an algorithm that learns the policy to maximize the return from the performance of the task, considering the convergence of trial and error.

A reward function is defined as an objective function for the environment in which the reward function values are often returned as integer or scalar variables associated with the states. A reward function can define an agent's goal in a given situation, defining what is useful or not useful for a particular object.

An action can be defined as a decision on an object to know how to make it, and a state can be defined as a factor to consider when a decision is made.

The action-based state may include a model of the surrounding environments, which may represent the previous state of the object peripheral conditions.

An object's policy is a mapping of possible states to a combination of all actions. An example of a stored policy mapping could be a lookup table.

A value function is a mapped object from a set of possible states that have a long-term reward evaluation that can be expected when the state is visited. The greedy policy is to choose a choice that leads to the state with the highest value.

As shown in FIG. 4, the Q-learning technique selects and applies an action after observing the state of the controller, and the system returns the state of the action to the controller. The controller then moves to the next state to receive the reward. The purpose of this learning is to retrieve a sequential order of actions that minimize the sum of future rewards.

An update algorithm for updating the auxiliary graph using the Q-learning technique is as follows.

Assuming that Q is a Q-table containing the Q values of each auxiliary graph of the network, each value in Q is composed of a set of residual energy of the node, a set of reliability of the node, and a time delay do.

The Q value of the auxiliary graph is updated with a positive or negative reward whenever data transfer is selected and the system manager learns the quality of the action that selects the auxiliary graph. After a certain period of time, the system manager examines the state of the Q values by the time it learns the values of the actions. Depending on the Q value, the supplemental graph is regenerated and replaced by a high performance supplemental graph by updating the values in the connection assisting graph algorithm (algorithm 1) described above.

If an application is scheduled to use the auxiliary graph, the acknowledgment of the last packet of the application flow includes a reward reflecting the network status of the auxiliary graph.

Here, the reward may include the expected residual energy of the nodes in the auxiliary graph, the reliability of the auxiliary graph, and the time delay due to the packet when the auxiliary graph is selected.

An authorized node in a part of the network constituting the auxiliary graph generated by the present invention may act as a malicious node by an attacker. That is, if a malicious node is selected for data transmission, it can promote a subsidiary graph by transmitting a positive reward even when no suitable performance is shown. Such malicious nodes cause performance degradation of applications scheduled to use auxiliary graphs.

Therefore, in order to prevent such a problem, it is possible to generate a subsidiary graph including a specific node by biasing a node selection decision so that a specific node with high reliability can be selected when the next node is selected in the process of generating the subsidiary graph have.

That is, when assigning weights to candidate nodes in the process of selecting a node, a bias margin is given to each of the weights assigned to a specific node so that a specific node can be selected, and then weights assigned to the bias margin are assigned to a specific node So that a specific node is selected in the node selection process according to the selection probability.

At this time, the weights to which the bias margin is given can be expressed as Equation (6) below.

&Quot; (6) "

Here, b means a predetermined bias margin.

In addition, a trust-based technique may be used to avoid generating false auxiliary graphs, and for this, a confidence evaluation function may be used that evaluates the received rewards for a particular auxiliary graph. The confidence evaluation function can evaluate the trust of the auxiliary graph by comparing the received reward for the auxiliary graph with the reward estimate for the auxiliary graph.

At this time, the trust value is given as the received value and can be determined based on a decision as to whether to accept or reject the received value.

Assuming that the empirical statistics on the auxiliary graph are assumed to be the expected reward of the auxiliary graph and the difference between the expected reward of the auxiliary graph and the received message is compared with the acceptable margin, If it is small, it is determined that the auxiliary graph is reliable. Otherwise, it is determined that the auxiliary graph is not reliable.

And, certain key bits can be used to prevent the data from changing during route computation. For example, by using two specific key bits to be transmitted from the transmitter to the receiver, these two key bits can be used to determine whether the transmitted data has changed.

As described above, in the auxiliary graph generation method according to the present invention, a predetermined bias margin is given to a weight of a specific node in order to select a reliable specific node as a node for generating a subsidiary graph, Thereby making it possible to improve the performance of the application scheduled to use the auxiliary graph.

In addition, the present invention can generate the auxiliary graph with high reliability by evaluating the reliability of the auxiliary graph, thereby reducing the energy cost of the entire network and improving the efficiency of the network.

FIG. 5 shows an example of an auxiliary graph generation time according to the number of nodes to which the present invention is applied. As shown in FIG. 5, when the present invention is applied (graph allocation, breadth traversal, and depth traversal) It can be seen that the auxiliary graph generation time is faster than the auxiliary graph generation time by the conventional method (instant graph routing). In particular, it can be seen that the generation time of the breadth traversal and the depth traversal of the auxiliary graph according to the present invention is faster than 2 seconds even when the number of nodes is 10,000. Thus, .

FIG. 6 shows an example of consumption of memory resources that is different from the number of nodes to which the present invention is applied. Graphs (graph allocation memory consumption, breadth traversal memory consumption) are graphs to which the present invention is applied. The memory resource consumes up to 140 to 145 [Mb], whereas if not using the instant graph routing, the memory resource consumes about 300 [Mb], so that the method according to the present invention consumes less memory resources Able to know.

FIG. 7 shows an example of generation time according to the auxiliary graph size according to the present invention. When the connection subgraph generation algorithm of the present invention is applied, the generation time according to the auxiliary graph size is constant regardless of the auxiliary graph size While the generation time according to the auxiliary graph size by the conventional method (instant graph routing) tends to increase as the auxiliary graph size increases.

FIG. 8 shows an example of the end-to-end delay time according to the number of event samples to which the present invention is applied, and shows an end to end time delay of a single application communication stream between a source node and a destination node will be.

FIG. 8 is a graph illustrating a result of a simple short-time delay according to the present invention (a proposed connection subgraph generation algorithm), which is calculated for a single application communication stream transmitting packets from a source node to a destination node in a 20- It can be seen that it is very small compared with the short time delay according to the conventional method (instant graph routing (ELHFR)). It can be seen that the conventional method has a maximum maximum time delay of 0.6 seconds and 0.7 seconds, whereas the present invention has a maximum maximum time delay of less than 0.3 seconds.

FIG. 9 shows an example of the energy analysis according to the number of subsidiary graphs to which the present invention is applied. In the present invention, the proposed energy aware subgraph generation method and the conventional auxiliary graph generation method that does not consider the energy cost (Without energy aware subgraph generation).

As shown in FIG. 9, the energy cost according to the conventional method has a constant energy cost regardless of the number of auxiliary graphs, whereas the present invention has a higher energy cost than the conventional method when the number of auxiliary graphs is 7 or less, As the number of supplementary graphs increases, the energy cost decreases significantly. That is, in a network having a number of auxiliary graphs above a certain level, the present invention can significantly reduce the energy cost.

FIG. 10 illustrates a service quality-based auxiliary graph generation system according to an exemplary embodiment of the present invention. FIG. 10 illustrates a service quality based auxiliary graph generation method according to the present invention.

10, a system 200 according to the present invention includes a statistical data collection unit 210, a service quality metric calculation unit 220, a selection probability calculation unit 230, a node selection unit 240, And a generating unit 250.

The statistical data collection unit 210 collects statistical data including quality of service (QoS) from the nodes of the network.

At this time, the statistical data collection unit 210 includes a list of costs for the paths between the source node and the destination node, a list of time delays for the paths, and a list of slots by the sum of links between the two nodes Statistical data may be collected and statistical data may be collected that includes the estimated energy costs of the packet transmission at each of the nodes, the expected packet transmission time, the reliability of the packet transmission, and the residual energy of each of the nodes .

The service quality metric calculation unit 220 calculates a service quality metric for generating an auxiliary graph using the collected statistical data.

The selection probability calculation unit 230 calculates the selection probability from the previously selected first node to each of the nodes in consideration of the energy cost for the nodes included in the service quality metric.

In this case, the selection probability calculator 230 may calculate the selection probability from the first node to each of the nodes by further considering the weight for the cost factor, the weight for the time delay factor, and the weight for the slot factor.

In this case, the selection probability calculator 230 arranges the weight factors for the cost factors, the weight factors for the time delay factors, and the weight factors for the slot factors in the candidate nodes for the first one of the nodes, It is possible to calculate the selection probability from the first node to each of the candidate nodes considering the energy cost.

In addition, the selection probability calculation unit 230 may be configured to calculate a weight for a cost factor, a time delay factor for a time delay factor, and so on so that a specific one of the candidate nodes for the first node can be selected as a next node, It is possible to calculate a selection probability for candidate nodes including a specific node by giving a predetermined bias margin to each of the weights for the weight and the slot factor and arranging the weights to which the bias margin is given to the specific node.

The node selecting unit 240 selects the second node, which is the next node, at the first node according to the calculated selection probability.

That is, the node selecting unit 240 selects a second node for generating the subsidiary graph among the candidate nodes for the first node.

The auxiliary graph generation unit 250 generates the auxiliary graph by performing the selection probability calculation by the selection probability calculation unit 230 and the node selection by the node selection unit 240 from the source node to the destination node.

At this time, the auxiliary graph generating unit 250 updates the statistical data using the Q-learning technique after a predetermined period from the time when the auxiliary graph is generated and the auxiliary graph is generated, and then, using the updated statistical data, Regenerated. Of course, the auxiliary graph generating unit 250 may control the statistical data collecting unit 210, the service quality metric calculating unit 220, the selection probability calculating unit 230, and the node selecting unit 240 after a predetermined period of time, However, the present invention is not limited to this, and it is also possible to control the configuration according to the present invention by checking the generation time of the auxiliary graph and the predetermined period in a separate control unit.

The service quality-based auxiliary graph generation method according to an exemplary embodiment of the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions recorded on the medium may be those specially designed and constructed for the present invention or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Magneto-optical media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

As described above, the present invention has been described with reference to particular embodiments, such as specific elements, and specific embodiments and drawings. However, it should be understood that the present invention is not limited to the above- And various modifications and changes may be made thereto by those skilled in the art to which the present invention pertains.

Accordingly, the spirit of the present invention should not be construed as being limited to the embodiments described, and all of the equivalents or equivalents of the claims, as well as the following claims, belong to the scope of the present invention .

Claims (15)

delete A method for generating an auxiliary graph from a source node to a destination node,
Collecting statistical data including quality of service (QoS) from nodes of the network;
Calculating a service quality metric for generating an auxiliary graph using the collected statistical data;
Calculating a selection probability from each of the nodes selected at the previously selected first node in consideration of the energy cost for the nodes included in the quality of service metric;
Selecting a second node as a next node in the first node according to the calculated selection probability; And
Calculating the selection probability and selecting the second node from the source node to the destination node to generate the auxiliary graph
Lt; / RTI >
The statistical data
A list of costs for the paths between the source node and the destination node, a list of time delays for the paths, and a list of slots by a sum of links between the two nodes. How to create auxiliary graph. A method for generating an auxiliary graph from a source node to a destination node,
Collecting statistical data including quality of service (QoS) from nodes of the network;
Calculating a service quality metric for generating an auxiliary graph using the collected statistical data;
Calculating a selection probability from each of the nodes selected at the previously selected first node in consideration of the energy cost for the nodes included in the quality of service metric;
Selecting a second node as a next node in the first node according to the calculated selection probability; And
Calculating the selection probability and selecting the second node from the source node to the destination node to generate the auxiliary graph
Lt; / RTI >
The statistical data
A packet transmission expected energy cost, a packet transmission expected time, a packet transmission expected reliability, and a residual energy of the corresponding node in each of the nodes. A method for generating an auxiliary graph from a source node to a destination node,
Collecting statistical data including quality of service (QoS) from nodes of the network;
Calculating a service quality metric for generating an auxiliary graph using the collected statistical data;
Calculating a selection probability from each of the nodes selected at the previously selected first node in consideration of the energy cost for the nodes included in the quality of service metric;
Selecting a second node as a next node in the first node according to the calculated selection probability; And
Calculating the selection probability and selecting the second node from the source node to the destination node to generate the auxiliary graph
Lt; / RTI >
The step of calculating the selection probability
Calculating a selection probability from the first node to each of the nodes by considering the weight for the cost factor, the weight for the time delay factor, and the weight for the slot factor, . 5. The method of claim 4,
The step of calculating the selection probability
Allocating a weight for the cost factor, a weight for the time delay factor, and a weight for the slot factor to candidate nodes for the first node among the nodes, and considering the allocated weights and the energy cost And calculating a selection probability from the first node to each of the candidate nodes. 6. The method of claim 5,
The step of calculating the selection probability
A predetermined bias margin is applied to each of a weight for the cost factor, a weight for the time delay factor and a weight for the slot factor so that a specific node among the candidate nodes can be selected as the second node, And assigning the weights with the bias margin to the specific node. A method for generating an auxiliary graph from a source node to a destination node,
Collecting statistical data including quality of service (QoS) from nodes of the network;
Calculating a service quality metric for generating an auxiliary graph using the collected statistical data;
Calculating a selection probability from each of the nodes selected at the previously selected first node in consideration of the energy cost for the nodes included in the quality of service metric;
Selecting a second node as a next node in the first node according to the calculated selection probability;
The step of calculating the selection probability and the step of selecting the second node from the source node to the destination node to generate the auxiliary graph; And
Updating the statistical data using a Q-learning technique after a predetermined period from the time when the auxiliary graph is generated, and regenerating the auxiliary graph using the updated statistical data
Generating an auxiliary graph based on quality of service.
A computer-readable recording medium having recorded thereon a program for executing the method according to any one of claims 2 to 7.
delete 10. An auxiliary graph generation system from a source node to a destination node,
A statistical data collecting unit for collecting statistical data including quality of service (QoS) from nodes of the network;
A service quality metric calculation unit for calculating a service quality metric for generating an auxiliary graph using the collected statistical data;
A selection probability calculation unit for calculating a selection probability from each of the nodes selected from the previously selected first node in consideration of an energy cost for the nodes included in the service quality metric;
A node selector for selecting a second node as a next node in the first node according to the calculated selection probability; And
An auxiliary graph generation unit for generating the auxiliary graph by performing the selection probability calculation by the selection probability calculation unit and the selection of the second node by the node selection unit from the source node to the destination node,
Lt; / RTI >
The statistical data
A list of costs for the paths between the source node and the destination node, a list of time delays for the paths, and a list of slots by a sum of links between the two nodes. Auxiliary graph generation system. 10. An auxiliary graph generation system from a source node to a destination node,
A statistical data collecting unit for collecting statistical data including quality of service (QoS) from nodes of the network;
A service quality metric calculation unit for calculating a service quality metric for generating an auxiliary graph using the collected statistical data;
A selection probability calculation unit for calculating a selection probability from each of the nodes selected from the previously selected first node in consideration of an energy cost for the nodes included in the service quality metric;
A node selector for selecting a second node as a next node in the first node according to the calculated selection probability; And
An auxiliary graph generation unit for generating the auxiliary graph by performing the selection probability calculation by the selection probability calculation unit and the selection of the second node by the node selection unit from the source node to the destination node,
Lt; / RTI >
The statistical data
A packet transmission expected energy cost, a packet transmission expected time, a packet transmission expected reliability, and a residual energy of the corresponding node in each of the nodes. 10. An auxiliary graph generation system from a source node to a destination node,
A statistical data collecting unit for collecting statistical data including quality of service (QoS) from nodes of the network;
A service quality metric calculation unit for calculating a service quality metric for generating an auxiliary graph using the collected statistical data;
A selection probability calculation unit for calculating a selection probability from each of the nodes selected from the previously selected first node in consideration of an energy cost for the nodes included in the service quality metric;
A node selector for selecting a second node as a next node in the first node according to the calculated selection probability; And
An auxiliary graph generation unit for generating the auxiliary graph by performing the selection probability calculation by the selection probability calculation unit and the selection of the second node by the node selection unit from the source node to the destination node,
Lt; / RTI >
The selection probability calculation unit
Wherein the selection probability for each of the nodes at the first node is calculated by further considering the weight for the cost factor, the weight for the time delay factor, and the weight for the slot factor. . 13. The method of claim 12,
The selection probability calculation unit
Allocating a weight for the cost factor, a weight for the time delay factor, and a weight for the slot factor to candidate nodes for the first node among the nodes, and considering the allocated weights and the energy cost And calculating a selection probability from the first node to each of the candidate nodes. 14. The method of claim 13,
The selection probability calculation unit
A predetermined bias margin is applied to each of a weight for the cost factor, a weight for the time delay factor and a weight for the slot factor so that a specific node among the candidate nodes can be selected as the second node, And assigning the weights with the bias margin to the specific node. 10. An auxiliary graph generation system from a source node to a destination node,
A statistical data collecting unit for collecting statistical data including quality of service (QoS) from nodes of the network;
A service quality metric calculation unit for calculating a service quality metric for generating an auxiliary graph using the collected statistical data;
A selection probability calculation unit for calculating a selection probability from each of the nodes selected from the previously selected first node in consideration of an energy cost for the nodes included in the service quality metric;
A node selector for selecting a second node as a next node in the first node according to the calculated selection probability; And
An auxiliary graph generation unit for generating the auxiliary graph by performing the selection probability calculation by the selection probability calculation unit and the selection of the second node by the node selection unit from the source node to the destination node,
Lt; / RTI >
The auxiliary graph generation unit
Wherein the statistical data is updated using a Q-learning technique after a predetermined period of time from the generation of the auxiliary graph, and the auxiliary graph is regenerated using the updated statistical data. Graph generation system.

Images