KR102120435B1 - Method for scheduling of wireless sensor network and controller of wireless sensor network - Google Patents

Abstract

A scheduling method in a wireless sensor network according to an aspect of the present invention includes constructing a BFS tree based on a width-first search based on a sink node; Dividing a node included in the BFS tree into a plurality of layers according to a hop distance from the sink node; Adding the sink node as a ruler node to the ruler set; A node of a third layer that is not adjacent to the first layer including the sink node is set as a candidate dominant node, and among the candidate dominant nodes, the node having the smallest delay when connecting to the sink node and the node are not adjacent to the node. Adding non-nodes to the set of rulers as ruler nodes; Selecting a linker node connecting the sink node and the ruler node through a shortest path among nodes of the second layer connecting the first layer and the third layer; Generating a ruler connection tree including ruler nodes of the ruler set and the connector node; Setting the dominant nodes not included in the dominant connection tree among the BFS trees as a transmitting node, and setting nodes included in the dominant connection tree as a backbone node; Allocating operation periods of the transmitting node, respectively, according to the dominant scheduling algorithm; And allocating an operation period of each backbone node so that data collision does not occur according to the backbone scheduling algorithm.

Description

Method for scheduling a wireless sensor network and a control device for the wireless sensor network that performs the same {METHOD FOR SCHEDULING OF WIRELESS SENSOR NETWORK AND CONTROLLER OF WIRELESS SENSOR NETWORK}

The present invention relates to a scheduling method of a wireless sensor network (WSN) operating in a duty cycle environment and a control device for a wireless sensor network performing the same.

With the development of wireless communication and electronics, small size, wireless short-range communication is possible, and low-power, multi-function and low-cost wireless sensors have been developed. In particular, recently, an Internet of Thing (IoT) sensor that connects to the Internet by embedding a sensor and a communication function in an object has been widely used.

Wireless sensor networks (WSNs) may be composed of nodes such as various wireless sensors and Internet of Things sensors. The wireless sensor network can monitor and monitor various status and environmental information such as temperature, humidity, illuminance, and pressure using a node without configuring a separate wired network. The wireless sensor network has recently been used for monitoring environmental conditions, structures and buildings, monitoring water and sewage networks, and chemical process and energy plant anomalies.

In such a wireless sensor network, nodes use limited resources. Therefore, nodes have limitations in the use of resources such as processing, storage, and battery. To solve this problem, a wireless sensor network technique using a duty-cycle has been proposed. In the duty cycle wireless sensor network, each node can periodically switch between an active mode and an inactive mode according to a time slot. In addition, the node may receive a data packet transmitted by another node in the activation mode. At this time, the node may broadcast data packets to other connected nodes regardless of the time slot.

Meanwhile, data aggregation is one of the most essential tasks in a wireless sensor network (WSN) where data from all sensor nodes is collected at a sink node. The WSN adopting the duty cycle technique makes the delay data aggregation problem more complicated due to the periodic dormant period of the sensor node.

The present invention proposes a new data set method that minimizes data set delay in duty cycled WSN.

The minimum time set scheduling problem necessary to implement WSN's real-time capabilities has been extensively studied. Several approaches have been proposed to reduce the latency of data aggregation in duty cycled WSN.

The scheduling algorithm (SA) is one of the earliest approaches and consists of two steps. In the first step, the Layered Structure Construction (LSC) algorithm constructs the connected dominating set (CDS) of the duty cycled WSN based on the Maximum Independent Set (MIS). Nodes not included in the CDS become the rulers of the network. The algorithm then builds a layered virtual backbone for data aggregation based on the CDS. However, the LSC algorithm does not take into account the delay due to the periodic power saving mode of the network. Thus, in the worst case, it becomes the backbone with the longest path delay. In the second step, the Working Period Scheduling (WPS) algorithm first collects data of all rulers to the rulers of the CDS. Data is then transmitted layer by layer from the ruler through the backbone. To avoid collisions, the algorithm delays the transmission of all nodes in the layer until all nodes in the lower layer finish transmission. This causes additional delay for all transmissions and makes the total aggregation delay of the SA method high.

Consider an aggregation process that collects data from all nodes of the duty cycled WSN to the sink node. Each sensor node in the network sends a single packet containing its own data and data received from neighbor nodes through the multi-hop path to the sink node. Complete data aggregation functions such as COUNT, MIN, MAX, SUM and AVERAGE are deployed to ensure that all packets have the same size. The aggregation schedule allocates an operation period and a reception node to all nodes except the sink to transmit data packets. Scheduled transmissions should be free of collisions while the allocated maximum duration of operation is minimized.

1 is a view for explaining a collision process occurring in the data aggregation process.

The letters and numbers displayed on each node indicate the identifier and active slot of each node. Each operating period consists of four time slots (ie, duty cycle is 25%). In the figure, a slot shown in black indicates an active slot, and a slot shown in white indicates an idle state. Data collected from each node is aggregated to sink node (a). However, as shown in the figure, the node d and the node e transmit data according to the active slots of the node b, and data collision occurs when the operation periods coincide. Likewise, the node e and the node f transmit data according to the active slot of the node c, and data collision occurs when the operation periods coincide.

Republic of Korea Patent Publication No. 10-2005-0010354 (name of invention: method for transmitting data with minimum power in a wireless sensor network)

One embodiment of the present invention is to solve the problems of the prior art described above, and is to provide a scheduling method of a wireless sensor network capable of minimizing data set delay in a duty cycled WSN and a control device for performing the same.

However, the technical problem to be achieved by the present embodiment is not limited to the technical problem as described above, and other technical problems may exist.

As a technical means for achieving the above technical problem, a scheduling method in a wireless sensor network according to an aspect of the present invention configures a BFS tree based on a width-first search based on a sink node for the wireless sensor network. step; Dividing a node included in the BFS tree into a plurality of layers according to a hop distance from the sink node; Adding the sink node as a ruler node to the ruler set; A node of a third layer that is not adjacent to the first layer including the sink node is set as a candidate dominant node, and among the candidate dominant nodes, the node having the smallest delay when connecting to the sink node and the node are not adjacent to the node. Adding non-nodes to the set of rulers as ruler nodes; Selecting a node that connects the sink node and the ruler node in the shortest path among nodes of the second layer connecting the first layer and the third layer as a connector node; Generating a ruler connection tree including ruler nodes of the ruler set and the connector node; Setting the dominant nodes not included in the dominant connection tree among the BFS trees as a transmitting node, and setting nodes included in the dominant connection tree as a backbone node; Scheduling data transmission between the transmitting node and the backbone nodes based on information about each active slot, and allocating operation periods of the transmitting node, respectively, according to the dominant scheduling algorithm; And an operation period of each backbone node to prevent data collision from occurring based on information on an operation period of backbone nodes that overhear data among neighboring nodes of the transmitting node to which the operation period is allocated according to a backbone scheduling algorithm. And assigning.

In addition, the control device of the wireless sensor network according to another embodiment of the present invention, a memory for storing a scheduling program of the wireless sensor network, a communication module connected to the wireless sensor network, and for executing a program stored in the memory Includes a processor. At this time, the processor configures a BFS tree based on a width-first search based on a sink node for the wireless sensor network by executing the program, and is included in the BFS tree according to a hop distance from the sink node. A node is divided into a plurality of layers, the sink node is added to a set of rulers as a ruler node, a node of a third layer not adjacent to the first layer including the sink node is set as a candidate ruler node, and a candidate ruler Among the nodes, a node having the smallest delay and a node not adjacent to the corresponding node are added to the ruler set when connecting to the sink node, and the second layer connecting the first layer and the third layer is added. A node that connects the sink node and the ruler nodes through the shortest path among nodes is selected as a connector node, and a ruler connection tree including the ruler nodes of the ruler set and the connector node is generated, and among the BFS trees, Set the dominant nodes not included in the dominant connection tree as the transmitting node, set the nodes included in the dominant connection tree as the backbone node, and according to the dominant scheduling algorithm, based on information about each active slot, the transmitting node and Operations of backbone nodes that overlay data among neighboring nodes of the transmission node to which the operation period is allocated, by scheduling data transmission between the backbone nodes, allocating operation periods of the transmission nodes, and according to a backbone scheduling algorithm. Based on the information on the period, an operation period of each backbone node is allocated so that data collision does not occur.

According to any one of the above-described problem solving means of the present invention, a ruler connection tree (CDS tree) in consideration of delay between nodes of a duty-cycled WSN through a delay-aware tree construction (DTC) algorithm is created and efficient data aggregation is performed in a network. For providing a virtual backbone. In addition, it is possible to perform a fast collision-free data collection schedule based on a backbone configured according to the initial fit aggregation scheduling (FAS).

1 is a view for explaining a collision process occurring in the data aggregation process.
2 is a view showing the configuration of a control device of a wireless sensor network according to an embodiment of the present invention.
3 is a flowchart illustrating a method of scheduling a wireless sensor network according to an embodiment of the present invention.
4 shows a pseudo code of a delay recognition tree construction algorithm according to an embodiment of the present invention.
5 illustrates an example in which a delay recognition tree construction algorithm is applied according to an embodiment of the present invention.
6 shows a pseudo code of a subject scheduling algorithm according to an embodiment of the present invention.
7 illustrates an example of a scheduling algorithm for a subject according to an embodiment of the present invention.
8 shows a pseudo code of a backbone scheduling algorithm according to an embodiment of the present invention.
9 shows an example of a backbone scheduling algorithm according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art to which the present invention pertains may easily practice. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. In order to clearly describe the present invention, parts not related to the description are omitted in the drawings, and like reference numerals are given to similar parts throughout the specification. Also, with reference to the drawings, even in the case of a configuration indicated by the same name, the drawing number may vary depending on the drawing, and the drawing number is for convenience of description only, and the concept, feature, and function of each configuration are indicated by the corresponding drawing number. Or, the effect is not to be interpreted limitedly.

Throughout the specification, when a part is "connected" to another part, this includes not only "directly connected" but also "electrically connected" with another element in between. . Also, when a part is said to “include” a certain component, it means that the component may further include other components, not exclude other components, unless otherwise stated. It should be understood that features or numbers, steps, actions, components, parts or combinations thereof are not excluded in advance.

In the present specification, the term “unit” or “module” includes a unit realized by hardware or software, and a unit realized by both, and one unit is realized by using two or more hardware. It may be, or two or more units may be realized by one hardware.

This technique consists of two stages: a layered network structure construction stage and a collision-free scheduling stage. In the first stage, the Delay-aware tree construction (DTC) stage, a CDS tree is created that takes into account the delay between nodes in the duty cycled WSN and prepares a virtual backbone for efficient data aggregation in the network. The second step, First-fit Aggregation Scheduling (FAS), performs a fast collision-free data collection schedule based on the configured backbone.

2 is a view showing the configuration of a control device of a wireless sensor network according to an embodiment of the present invention.

The control device 100 of the wireless sensor network includes a communication module 110, a memory 120, a processor 130, and a database 140.

The communication module 110 maintains a connection state with a wireless sensor network forming a network by connecting a plurality of wireless sensor nodes to each other, and performs data communication. The communication module 110 may be a device including hardware and software necessary for transmitting and receiving a signal such as a control signal or a data signal through a wired or wireless connection with another network device.

In the memory 120, a scheduling program of a wireless sensor network is recorded. In addition, it performs a function of temporarily or permanently storing data processed by the processor 130. Here, the memory 120 may include volatile storage media or non-volatile storage media, but the scope of the present invention is not limited thereto.

The processor 130 controls the entire process in which the control device of the wireless sensor network performs the scheduling program of the wireless sensor network. Here, the processor 130 may include all kinds of devices capable of processing data, such as a processor. Here, a'processor' may mean a data processing device embedded in hardware having physically structured circuits, for example, to perform functions represented by codes or instructions included in a program. As an example of such a data processing device embedded in hardware, a microprocessor, a central processing unit (CPU), a processor core, a multiprocessor, and an application-specific integrated ASIC circuit, a field programmable gate array (FPGA), or the like, but the scope of the present invention is not limited thereto.

The processor 130 configures a BFS tree based on a width priority search based on a sink node for the wireless sensor network by executing a scheduling program of the wireless sensor network, and according to a hop distance from the sink node, A node included in the BFS tree is divided into a plurality of layers, the sink node is added to the set of rulers as a ruler node, and a node of a third layer that is not adjacent to the first layer including the sink node is a candidate ruler node. Set, add the node with the smallest delay and the node not adjacent to the node to the set of rulers when connecting with the sink node among candidate ruler nodes, and connect the first layer and the third layer A node that connects the sink node and the ruler nodes in the shortest path among nodes of the second layer is selected as a connector node, and generates a ruler connection tree including the ruler nodes of the ruler set and the connector nodes, Based on information about each active slot, among the BFS trees, set the dominant nodes not included in the dominant connection tree as a transmitting node, set the nodes included in the dominant connection tree as a backbone node, and according to the dominant scheduling algorithm By scheduling data transmission between the transmitting node and the backbone nodes, each of the operating periods of the transmitting node is allocated, and according to a backbone scheduling algorithm, data of neighbor nodes of the transmitting node to which the operating period is allocated is overheard. Based on the information on the operation period of the backbone nodes, each operation of allocating the operation period of each backbone node is performed so that data collision does not occur.

The database 140 stores information about individual nodes constituting the wireless sensor network. In addition, data accumulated while performing the scheduling program of the wireless sensor network is stored.

3 is a flowchart illustrating a method of scheduling a wireless sensor network according to an embodiment of the present invention.

First, a ruler connection tree is generated according to a delay recognition tree construction algorithm (S310).

To this end, as described in FIGS. 4 and 5 below, a BFS tree is constructed based on a width-first search based on a sink node for a wireless sensor network. Then, nodes included in the BFS tree are divided into a plurality of layers according to the hop distance from the sink node. Basically, the sink node is set up as a dominant node, and it is added to the dominant set. The layer including the sink node is set as the first layer, and the layers below it are divided into a second layer, a third layer, and the like.

The second layer adjacent to the first layer is excluded from the ruler candidate, and a node of the third layer not adjacent to the first layer is set as a candidate ruler node. In addition, when connecting to the sink node among the candidate dominant nodes, the node having the smallest delay and the node not adjacent to the node are added to the dominant set as the dominant node.

Then, among the nodes of the second layer connecting the first layer and the third layer, a node connecting the sink node and the dominant nodes in the shortest path is selected as the connector node. In this way, a ruler connection tree including ruler nodes and connector nodes of the ruler set is generated.

The process of generating the ruler connection tree will be described in more detail with reference to FIGS. 4 and 5 below.

4 illustrates a pseudo code of a delay recognition tree construction algorithm according to an embodiment of the present invention.

As can be seen in Figure 4, the DTC algorithm consists of three steps.

)의 싱크 노드 (

)를 베이스로 하는 넓이 우선 탐색(breadth first search, BFS)에 기반한 BFS 트리를 구성한다. 네트워크의 모든 노드는 싱크노드(s)와의 홉 거리에 따라 계층 (

및

)이 구분되는데,

는 BFS 트리에서 가장 큰 계층을 나타낸다.

에는 싱크 노드만 포함된다. 싱크는 처음에 지배자로 선택되어 지배자 집합(Dominating Set, DS)에 추가된다. Network in the first stage (

) Sink node (

Constructs a BFS tree based on breadth first search (BFS) based on ). All nodes in the network are hierarchical according to the hop distance from the sync node (s) (

And

) Is separated,

Indicates the largest layer in the BFS tree.

Includes only sink nodes. The sink is initially selected as the ruler and added to the Dominating Set (DS).

까지 두번째 단계와 세번째 단계의 여러 반복을 실행하여 싱크노드(s)를 바탕으로 하는 CDS 트리를 구성한다. 지연 인식 트리 구조를 용이하게 만들기 위해 DTC 알고리즘은 다음 수학식 1과 같이 노드 쌍

사이의 휴지 지연(sleep delay)을 결정한다.The algorithm is layered at layer 1

Until this time, multiple iterations of the second and third steps are executed to construct a CDS tree based on the sync node (s). To facilitate the delay-aware tree structure, the DTC algorithm uses node pairs as shown in Equation 1 below.

Determine the sleep delay between.

[Equation 1]

는 노드 u가 노드 v의 활성 시간 슬롯에서 데이터를 노드 v로 전송하기 위해 대기해야하는 지연을 나타낸다. 경로의 휴지 지연은 경로의 모든 에지의 누적 지연 시간이다.

Denotes the delay that node u must wait to transmit data to node v in the active time slot of node v. The idle delay of the path is the cumulative delay time of all edges of the path.

를 계층(i)의 노드들의 집합이라고 한다. 두번째 단계에서 어떠한 지배자와도 인접하지 않은

의 노드는 후보 지배자가 되고 연결 지배자(CD)를 설정하기 위해 추가된다(5번 라인). 후보 지배자로부터 가장 가까운 지배자를 식별하기 위해, 후보 지배자로부터 2홉 거리 내의 모든 지배자까지의 지연 측면으로 최단 경로가 확립된다(7번 라인). 이러한 지배자는 항상 네트워크 연결성 때문에 상위 계층 j(

)에 존재한다. 후보 지배자에서 가장 가까운 지배자까지의 최단 경로와 경로의 휴지 지연은 추후 사용을 위해 기록된다 (8-9번 라인).

는 후보 지배자 u와 가장 가까운 지배자 v사이의 지연에 대한 최단 경로를 나타낸다.

의 휴지 지연은 수학식 2와 같다.

Is referred to as a set of nodes in layer (i). In the second stage, it is not adjacent to any ruler

The node of is added to become a candidate ruler and establish a connection ruler (CD) (line 5). To identify the closest ruler from the candidate ruler, the shortest path is established in terms of delay from the candidate ruler to all rulers within two hops (line 7). These rulers are always on top layer j( due to network connectivity.

). The shortest route from the candidate ruler to the closest ruler and the pause delay of the route are recorded for future use (lines 8-9).

Denotes the shortest path to the delay between the candidate ruler u and the closest ruler v.

The pause delay of is equal to Equation 2.

[Equation 2]

는 경로의 중간 노드이다. here

Is the middle node of the path.

) 를 생성한다. 처음에는 트리에 싱크 노드 s 만 포함된다. 각 계층에서 연결 지배자(CD)에 속한 노드(u)는 후보 지배자 집합의 모든 노드 중에서

의 값이 가장 작은 경우 지배자로 선택된다. 선택된 지배자와 모든 이웃은 집합에서 제거된다 (11-12번 라인). 노드(u)와 경로

의 중간 노드 (

)는 CDS 트리(

)에 추가된다 (13-14번 라인). 노드(

)는 연결자로 불리며 트리에서 노드(u)의 부모 노드(

)가 된다. 노드(u)의 가장 가까운 지배자(v)가 이미 CDS 트리(

)에 있고

가 된다는 것은 주목할 가치가 있다.The DTC algorithm selects a ruler from among candidates, and in the third stage, a CDS tree (

). Initially, the tree contains only the sink node s. The nodes (u) belonging to the connection ruler (CD) in each layer are among all nodes in the candidate ruler set.

If the value of is the smallest, it is selected as the ruler. The selected ruler and all neighbors are removed from the set (lines 11-12). Node (u) and path

Middle node (

) Is the CDS tree (

) (Lines 13-14). Node(

) Is called a connector and is the parent node of node u in the tree (

). The closest ruler (v) of node (u) is already the CDS tree (

)

It is worth noting that it becomes.

)까지 반복한다. 그런 다음 계층 집합과 최종 계층을 업데이터하여 업데이트하여 모든 노드가 CDS 트리(

)에서 상위 노드의 인접한 하위 레이어에 있도록 한다 (15번 라인). 본 발명은 CDS 트리에 없는 노드를 지배자로 지정한다.The DTC algorithm divides the second and third steps into the final layer (

Repeat until ). Then, update the hierarchy set and the final hierarchy to update all nodes so that the CDS tree (

) In the adjacent lower layer of the upper node (line 15). The present invention designates nodes that are not in the CDS tree as rulers.

5 illustrates an example in which a delay recognition tree construction algorithm is applied according to an embodiment of the present invention.

The illustrated WSN network includes a total of 13 nodes, and the duty cycle is set to 25% (τ=4). As shown in (a), initially, it is divided into four layers (first layer to fourth layer). The sink node (a) is selected as the ruler and is included in the CDS tree. Nodes of the second layer are adjacent to the sink node and thus become dominant. Nodes of the third tier become candidate rulers. The delay of all paths from the candidate ruler to the sink node (a), which is the second-hop higher ruler, is shown in (b). Each delay is determined based on the active slot of each node. For example, since the active slot of node (a) is 3 and the active slot of node (b) is 2, data transmitted while node (b) is active is transmitted to node (a) after a delay of 1 do. Likewise, since the active slot of node (a) is 3 and the active slot of node (d) is 0, data transmitted in the state where node (d) is activated is transmitted to node (a) after a delay of 4. In this way, it is possible to check the delay in the path between each node based on the active slot of each node.

), 최초로 지배자로 선택된다. 제 3 계층의 노드(e, g)는 지배자(f)와 인접하고 있으므로 피지배자가 된다. 마찬가지 방법으로, 노드(h)도 지배자로 선택된다.As shown in (c), the DTC algorithm determines the minimum path in terms of delay, and calculates the delay for the closest upper ruler in each node of the third layer. Node (f) has the smallest delay to the parent ruler (

), first chosen as ruler. The nodes (e, g) of the third layer are adjacent to the ruler (f), and thus become the dominant. In the same way, node h is also selected as the ruler.

Similarly, as shown in (d), nodes f and h are connected to the CDS tree through two connecting nodes b and d, respectively.

As shown in (e), the DTC algorithm is also performed on the node included in the fourth layer. That is, when the fourth layer is further included as a lower layer of the third layer, the node of the fourth layer is set as a candidate dominant node, and among the candidate dominant nodes, the delay is the greatest when connecting with the dominant node of the third layer. A small node and a node not adjacent to the node are added to the ruler set as the ruler node. By repeating this process, finally, as shown in (f), the hierarchical sets are updated based on the CDS tree. The hierarchies of the nodes g, i, l increase significantly more than the hierarchies of their parent nodes.

Referring to FIG. 3 again, an operation period of the transmitting node is allocated according to the dominant scheduling algorithm (S320).

First, among the BFS trees configured above, dominant nodes not included in the dominant connection tree are set as the transmission node, and nodes included in the dominant connection tree are set as the backbone node. Then, according to the dominant scheduling algorithm, data transmission between a transmitting node and the backbone nodes is scheduled based on information on each active slot, and an operation period of the transmitting node is respectively allocated.

In particular, when allocating the operation period of the transmitting node, the backbone nodes are classified according to the order of every active slot to generate a plurality of backbone node sets (b in FIG. 7 ), and data is transmitted to the backbone node set for each active slot The minimum set of covers connected to all of the transmitting nodes is specified in the set of backbone nodes (c, d, e in FIG. 7). Then, for each active slot, a first operation period is allocated to a task in which the transmitting node transmits data to the backbone node included in the minimum cover set (c, d, e of FIG. 7 ).

At this time, when a plurality of transmitting nodes are in an exclusive neighbor relationship to the backbone node, a first operation period is selected for the operation of randomly selecting one of the transmitting nodes and transmitting data to the backbone node by the selected transmitting node. Can be assigned.

6 and 7 below, the subject scheduling algorithm will be described in more detail.

6 shows a pseudo code of a subject scheduling algorithm according to an embodiment of the present invention.

)를 참조하면, FAS 방식은 네트워크 내의 모든 노드들로부터 싱크 노드까지 데이터를 수집하는 효율적인 스케줄을 구성할 수 있다. 최초 적합 집계 스케줄링 방법은 두 단계로 구성된다. CDN tree of WSN as aggregate backbone (

), the FAS method can configure an efficient schedule for collecting data from all nodes in the network to the sink node. The initial adaptive aggregation scheduling method consists of two steps.

First, the data of all dominants is collected as a node of the backbone, and then the data of the backbone node is aggregated to the sink node without collision in a bottom-up manner. The backbone node contains all the rulers and connectors selected in the DTC algorithm.

)에 따르면 네트워크의 모든 노드들은 두 개의 분리된 집합으로 구분되는데, 백본의 노드 집합(R)과 피지배자 집합(S)으로 구분된다 (1번 라인). 피지배자 집합(S)내의 각각의 피지배자는 백본 노드 집합(R)내의 적어도 하나의 백본 노드, 즉 그 지배자에 인접해있다. 동작 기간 1부터 시작하여 알고리즘은 피지배자 집합(S) 내의 각 송신 노드에 대해 동작 기간을 할당하여 데이터 패킷이 백본 노드 집합(R)의 수신 노드로 충돌 없이 전송될 수 있도록 한다. 할당된 동작 기간은 최소 지연 집합을 가능하게 하기 위해 가능한 작아야 한다.CDS tree(

According to ), all nodes in the network are divided into two separate sets, the node set (R) of the backbone and the set of subjects (S) (line 1). Each dominant in the dominant set S is adjacent to at least one backbone node in the set of backbone nodes R, i.e., its dominant. Starting from the operation period 1, the algorithm allocates an operation period to each transmitting node in the set of the dominant S so that data packets can be transmitted without collision to the receiving node of the backbone node set R. The assigned duration of operation should be as small as possible to enable a minimal set of delays.

(

)으로 그룹화 한다(3-4번 라인).

가

에 있는 노드들에 인접한 송신 노드들의 서브 세트를 표시한다고 하자 (7번 라인). 피지배자 스케줄링 알고리즘은

의 모든 노드를 커버할 수 있는 최소 커버 집합

(

)을 찾는다 (8번 라인). 최소 커버 집합의 특성에 따르면, 최소 커버 집합의 각 노드(

)는 해당 노드(u)에만 인접한 적어도 하나의 독점적인 이웃 노드(v,

)를 가져야한다(10번 라인). 만약, 수신 노드에 둘 이상의 독점적인 이웃이 있는 경우 알고리즘은 이웃 노드 중 하나를 송신 노드로 무작위로 선택한다. 이웃 노드(v,

)의 송신 스케줄을

로 나타내면(11번 라인), 노드 (v)가 제 m 동작 기간의 시간 슬롯(

) 에서 노드(u) 로 데이터 패킷을 보내도록 스케줄링됨을 의미한다. 각 활성 슬롯(i)의 이웃 노드들의 쌍(

)들이 존재한다. 노드들의 쌍들 사이의 전송은 충돌 없이 동일한 동작 기간 m에서 동시에 스케줄링 될 수 있다. To ensure collision-free data collection, the dominant scheduling allows the receiving node to have different subsets depending on the active slot.

(

) (Line 3-4).

end

Suppose that we indicate a subset of transmitting nodes adjacent to the nodes in (line 7). The dominant scheduling algorithm

Set of minimum covers that can cover all nodes in a

(

) (Line 8). According to the characteristics of the minimum cover set, each node of the minimum cover set (

) Is at least one exclusive neighbor node (v, only adjacent to the corresponding node u)

) (Line 10). If the receiving node has two or more exclusive neighbors, the algorithm randomly selects one of the neighboring nodes as the transmitting node. Neighbor node (v,

)

When represented by (line 11), node (v) is the time slot of the mth operation period (

) Means that it is scheduled to send a data packet to node u. Pair of neighboring nodes in each active slot (i)

) Are present. The transmission between pairs of nodes can be scheduled simultaneously in the same operating period m without collision.

)만이 동작 기간 m 에서 패킷을 수신하도록 보장되지만(13번 라인), 이웃 노드(

)들은 동일한 시간 슬롯에서 활성 상태이기 때문에 송신을 엿들을 수 있다. 피지배자 스케줄링 알고리즘은 이와 같은 엿듣게 되는 동작 기간을 기록하여, 인접 노드가 나중에 데이터 전송시에 간섭하지 못하도록 한다 (14-15번 라인). 남아있는 피지배자는 모든 피지배자가 스케줄링 될 때까지, 다음 활성 슬롯에서 동작 기간에 반복적으로 스케줄링 된다 (16번 라인). 상이한 활성 슬롯들에서의 전송들이 서로 간섭하지 않을 것이라는 점은 주목할 가치가 있다. When the dominant (v) is scheduled to transmit its own data packet, it is removed from the set of transmitting nodes (s) (lines 11-12). Corresponding receiving node (

) Is guaranteed to receive packets only during operation period m (line 13), but neighbor nodes (

) Can overhear the transmission because they are active in the same time slot. The dominant scheduling algorithm records such an overheard period of operation, preventing adjacent nodes from interfering with data transmission later (lines 14-15). The remaining dominant is repeatedly scheduled in the operation period in the next active slot, until all dominants are scheduled (line 16). It is worth noting that transmissions in different active slots will not interfere with each other.

7 illustrates an example of a scheduling algorithm for a subject according to an embodiment of the present invention.

)에 따르면, 모든 노드들은 (b)에 도시된 바와 같이 송신 집합(S = {c, e, k,m})과 수신 집합(R ={a, b, d, f , h, j, g, i, l})으로 구분된다. 피지배자 스케줄링 알고리즘은 각 활성 슬롯마다 송신 집합(S)의 노드로부터 수신 집합(R)의 노드로 송신(transmission)을 스케줄링한다. The CDS tree shown in (a) (

According to ), all nodes have a transmit set (S = {c, e, k,m}) and a receive set (R ={a, b, d, f, h, j, g, as shown in (b)). , i, l}). The dominant scheduling algorithm schedules transmission from a node of the transmission set S to a node of the reception set R for each active slot.

)만 송신 노드(

)로부터 데이터 패킷을 수신할 수 있다. 최소 커버 집합(

)이 송신 집합(

)의 모든 노드를 커버하도록 구성된다. 동일한 동작 기간에 스케줄링된 노드(c, m)로부터 데이터를 수신하는 노드(f, m)이 최소 커버 집합(

)에 포함된다. As shown in (c), in the active slot 0, the nodes of the receive set (

Only the sending node (

) To receive a data packet. Minimum cover set (

) Is the send set (

) To cover all nodes. A node (f, m) receiving data from the nodes (c, m) scheduled in the same operation period has a minimum set of covers (

).

Similarly, as shown in (d) and (e), the dominant scheduling algorithm transmits data from the remaining transmitting nodes (e, k) to the receiving nodes (i, g) in the first active slot and the second active slot, respectively. You can schedule the transmission.

)에 포함된 모든 피지배자가 데이터를 모두 전송하면 종료된다.As shown in (f), the process of collecting the subject data is the CDS tree (

) Is terminated when all the parties included in) transmit all the data.

Referring back to FIG. 3, an operation period of a backbone node is allocated according to a backbone scheduling algorithm (S330).

According to the backbone scheduling algorithm, the operation period of each backbone node is allocated so that data collision does not occur based on information on the operation period of the backbone nodes that overhear data among neighboring nodes of the transmitting node to which the operation period is assigned. .

In particular, information on the operation period of the backbone nodes that overhear data among neighbor nodes of the transmitting node is collected, and the operation period is started from the backbone nodes of the lowest layer based on the operation period information of the backbone nodes that overhear data. Assign. At this time, when allocating the operation period, the operation period of each backbone node is allocated so that data collision does not occur by referring to the operation period that neighbor nodes adjacent to the corresponding backbone node overhear (FIG. 9).

In particular, when the operation period is allocated, when the backbone node receiving data transmission from the transmission node to which the first operation period is assigned allocates an operation period for transmitting data to the parent node, the parent node is assigned the first operation period. When operating in an active slot that does not overhear one or more transmitting nodes, the backbone node allocates a first operation period to the operation of transmitting data to the parent node, and the parent node transmits one or more operations to which the first operation period is allocated When operating in an active slot listening to a node, the backbone node allocates a second operation period to a task of transmitting data to the parent node (FIG. 9E).

The backbone scheduling algorithm will be described in more detail with reference to FIGS. 8 and 9 below.

8 shows a pseudo code of a backbone scheduling algorithm according to an embodiment of the present invention.

)계층에서 부모 노드로 데이터 패킷을 전송하는 첫 번째로 가용한 동작 기간을 결정한다. 노드는 활성 상태에서만 데이터를 수신할 수 있기 때문에 노드(u)가 동작 기간에 있는 노드(p(u))이전에 활성화되어 있으면 마지막 수신 노드(rx(u))와 동일한 동작 기간에 있는 노드(p(u))로 패킷을 전송할 수 있다. 그렇지 않으면 노드(p(u))가 다음 동작 기간에서 활성화될 때까지 기다려야한다 (3-4번 라인). 노드가 두 번째 단계에서 자식 노드 뿐만 아니라 첫 번째 단계에서 피지배자로부터 모든 데이터를 받은 후에 데이터를 전송할 준비가 되었음을 주목할 필요가 있다. 피지배자로부터 어떠한 데이터도 받지 못한 잎(leaf) 노드는 동작 기간 1에 준비되어 있다.The second phase of the FAS system begins after collecting all data from the subject. 8 shows a pseudo code of a backbone scheduling algorithm for aggregating data of the ruler and the connector. This algorithm consists of a CDS tree consisting of all nodes (

)Determine the first available operation period for transmitting data packets from the layer to the parent node. Since the node can only receive data in the active state, the node in the same operating period as the last receiving node (rx(u)) if node u is active before the node in operation period (p(u)) p(u)). Otherwise, we have to wait for node p(u) to become active in the next period of operation (lines 3-4). It is worth noting that the node is ready to transmit data after receiving all data from the child node as well as the child in the first stage in the second stage. A leaf node that has not received any data from the dominant is ready for operation period 1.

)에 있는 이웃들이 다른 스케줄링 된 전송을 엿듣는 모든 동작 기간을 수집한다. 그것은 노드(u)의 금지된 동작 기간과 같은 동작 기간을 말하며

에 저장한다 (5-6번 라인). 이 알고리즘은 노드(u)가 동작 기간 동안 패킷을 전송하는 것을 방지하고 노드(u)에 대한 최소 전송 동작 기간을 검색한다 (7-8 번 라인). 마지막 수신 동작 기간(p(u))과 노드(u)의 CDS 트리(

)에서 활동 중인 모든 이웃 노드의 엿듣는 동작 기간(a(p((u)))은 충돌을 회피하기 위해 업데이트 된다. 백본 스케줄링 알고리즘은 모든 노드의 데이터가 싱크 노드에서 집계되도록 트리에서 각 노드에 대해 최초 적합 동작 기간을 상향식 방식으로 할당한다.Note that all neighboring nodes of node u that are activated in the same time slot as node p(u) can overhear data transmission from node u to the parent node. Data transmission from node u can interfere with other scheduled transmissions, causing collisions in neighboring nodes. To ensure a collision-free dataset, the backbone scheduling algorithm uses the node u's CDS tree (

) Collects all of the operation periods when neighbors in the eavesdrop on other scheduled transmissions. It refers to the period of operation equal to the forbidden period of operation of node u.

To the line (lines 5-6). This algorithm prevents node u from sending packets during the operation period and retrieves the minimum transmission operation period for node u (lines 7-8). CDS tree of last receive operation period (p(u)) and node (u) (

The period of eavesdropping (a(p((u)))) of all neighboring nodes active in) is updated to avoid collisions.The backbone scheduling algorithm is applied to each node in the tree so that data from all nodes is aggregated at the sink node. For the first time, an appropriate fit operation period is allocated in a bottom-up manner.

9 shows an example of a backbone scheduling algorithm according to an embodiment of the present invention.

The data transmission process of the subjects shown in FIG. 7(f) is shown again in FIG. 9(a). As shown in (b), the operation period overheard in the data transmission process is updated for the backbone node. For example, when node e transmits data, node i can receive data in the active slot 1 of the next operation period, and nodes b, j among the neighboring nodes of node e ) Is activated in the active slot 2, so data cannot be overheard, but node f is activated in the active slot 0, so that it can be overheard in the next operation period. In this way, for neighbor nodes of each dominant node, information about an operation period in which the dominant node overhears data when transmitting data is collected.

)들로부터 시작된다. 엿듣는 동작 기간은 노드(i)의 모든 이웃 노드들은 공백상태이므로, 금지된 집합은

이 된다. 노드(i)는 그 부모 노드(j)에 앞서 동작 기간에 활성화되므로, 노드(i)는 피지배자 노드(e)로부터 패킷을 수신할 때, 동일한 동작 기간(1)에 패킷을 노드(j)로 전송할 수 있다. 반면에, 노드(l)는 그 부모 노드(g)가 동작 기간(1)에 피지배자 노드(k)로부터 스케줄된 데이터 전송을 받는 동작 상태이므로, 동작 기간(2)로 데이터 전송을 지연시킨다. 제 4 계층 내지 제 1계층의 노드의 데이터 전송 스케줄은 (d) 내지 (f)에 각각 도시된다. 본 예시에서는 데이터 집계 스케줄이 동작 기간(3)에서 종료된다.As shown in (c), the backbone scheduling algorithm is a node of the fifth layer (

). The period of eavesdropping is blank for all neighbor nodes of node i, so the forbidden set

It becomes. Since the node i is activated during the operation period prior to its parent node j, the node i sends the packet to the node j in the same operation period 1 when receiving a packet from the dominant node e. Can transmit. On the other hand, the node l is an operation state in which the parent node g receives the scheduled data transmission from the dominant node k in the operation period 1, thereby delaying the data transmission to the operation period 2. Data transmission schedules of the nodes of the fourth layer to the first layer are shown in (d) to (f), respectively. In this example, the data aggregation schedule ends in the operation period (3).

The scheduling method of the wireless sensor network according to the embodiment of the present invention described above may also be implemented in the form of a recording medium including instructions executable by a computer, such as a program module executed by a computer. Such recording media include computer readable media, which can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media, removable and non-removable media. In addition, computer readable media includes computer storage media, which are volatile and nonvolatile implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. , Removable and non-removable media.

The above description of the present invention is for illustration only, and a person who is investigating ordinary knowledge in the technical field to which the present invention pertains can understand that it can be easily modified to other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as distributed may be implemented in a combined form.

Further, although the methods and systems of the present invention have been described in connection with specific embodiments, some or all of their components or operations may be implemented using a computer system having a general purpose hardware architecture.

The scope of the present invention is indicated by the following claims rather than the detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted to be included in the scope of the present invention. .

100: wireless sensor network control device 110: communication module
120: memory 130: processor

Claims (11)

In the scheduling method in a wireless sensor network,
Constructing a BFS tree based on a width-first search based on a sink node for the wireless sensor network;
Dividing a node included in the BFS tree into a plurality of layers according to a hop distance from the sink node;
Adding the sink node as a ruler node to the ruler set;
A node of a third layer that is not adjacent to the first layer including the sink node is set as a candidate dominant node, and among the candidate dominant nodes, the node having the smallest delay when connecting to the sink node and the node are not adjacent to the node. Adding non-nodes to the set of rulers as ruler nodes;
Selecting a node that connects the sink node and the ruler node in the shortest path among nodes of the second layer connecting the first layer and the third layer as a connector node;
Generating a ruler connection tree including ruler nodes of the ruler set and the connector node;
Setting the dominant nodes not included in the dominant connection tree among the BFS trees as a transmitting node, and setting nodes included in the dominant connection tree as a backbone node;
Scheduling data transmission between the transmitting node and the backbone nodes based on information about each active slot, and allocating operation periods of the transmitting node, respectively, according to the dominant scheduling algorithm; And
According to the backbone scheduling algorithm, the operation period of each backbone node is allocated so that data collision does not occur based on information on the operation period of the backbone nodes that overhear data among neighbor nodes of the transmitting node to which the operation period is assigned. Scheduling method in a wireless sensor network comprising the step of. According to claim 1,
When the BFS tree further includes a fourth layer as a lower layer of the third layer,
The node of the fourth layer is set as a candidate ruler node, and among the candidate ruler nodes, a node having the smallest delay and a node not adjacent to the node when connecting to the ruler node of the third layer is the ruler node. The scheduling method in the wireless sensor network is further performed before the step of adding the set to the set is performed before generating the ruler connection tree. According to claim 1,
The step of allocating the operation period of the transmitting node according to the dominant scheduling algorithm is
Classifying the backbone nodes according to the order of every active slot to generate a plurality of backbone node sets;
Specifying, in the backbone node set, a minimum cover set connected to all transmitting nodes that transmit data to the backbone node set for each active slot; And
A scheduling method in a wireless sensor network in which a first operation period is allocated to a task in which the transmitting node transmits data to a backbone node included in the minimum cover set for each active slot. The method of claim 3,
When a plurality of transmitting nodes in the backbone node are in an exclusive neighbor relationship, the first operation period is selected to randomly select one of the transmitting nodes and the selected transmitting node transmits data to the backbone node. Scheduling method in a wireless sensor network that allocates. According to claim 1,
According to the backbone scheduling algorithm, the step of allocating the operation period of each backbone node is
Collecting information about an operation period of backbone nodes that overhear data among neighboring nodes of the transmitting node;
And allocating an operation period from backbone nodes of the lowest layer based on the operation period information of the backbone nodes overheard the data,
When allocating the operation period of the backbone node, the operation method of each backbone node is allocated so that data collision does not occur by referring to the operation period that neighbor nodes adjacent to the corresponding backbone node overhear. The method of claim 5,
When allocating the operation period of the backbone node, when the backbone node receiving data transmission from the transmission node to which the first operation period is assigned allocates an operation period for transmitting data to its parent node,
When the parent node operates in an active slot that does not overhear one or more transmitting nodes to which the first operation period is assigned, the backbone node allocates a first operation period to the operation of transmitting data to the parent node,
When the parent node operates in an active slot that eavesdrops on one or more transmitting nodes to which the first operation period is assigned, the wireless sensor assigns a second operation period to the operation of transmitting data to the parent node by the backbone node. Scheduling method in the network. In the control device of the wireless sensor network,
A memory in which the scheduling program of the wireless sensor network is stored,
A communication module connected to the wireless sensor network, and
It includes a processor for executing a program stored in the memory,
The processor configures a BFS tree based on a width priority search based on a sink node for the wireless sensor network by executing the program, and a node included in the BFS tree according to a hop distance from the sink node Is divided into a plurality of layers, the sink node is added to a set of rulers as a ruler node, a node of a third layer not adjacent to the first layer including the sink node is set as a candidate ruler node, and a candidate ruler node When connecting to the sink node, a node having the lowest delay and a node not adjacent to the node are added to the set of rulers as a ruler node, and a node of a second layer connecting the first layer and the third layer Among them, a node connecting the sink node and the ruler nodes by the shortest path is selected as a connector node, and a ruler connection tree including the ruler nodes of the ruler set and the connector node is generated, and the ruler among the BFS trees Set the dominant nodes not included in the connection tree as the transmitting node, set the nodes included in the dominant connection tree as the backbone node, and according to the dominant scheduling algorithm, based on the information about each active slot, the transmitting node and the By scheduling data transmission between backbone nodes, each operation period of the transmission node is allocated, and according to a backbone scheduling algorithm, operation periods of backbone nodes that overhear data among neighbor nodes of the transmission node to which the operation period is allocated Based on the information on, the control device of the wireless sensor network is to allocate the operation period of each backbone node so that no data collision occurs. The method of claim 7,
When the processor allocates each operation period of the transmitting node according to the dominant scheduling algorithm, the processor classifies the backbone nodes according to the order of every active slot, generates a plurality of backbone node sets, and sets the backbone node for each active slot The minimum cover set connected to all the transmission nodes transmitting data to the backbone node set is specified, and for each active slot, the transmission node transmits data to the backbone node included in the minimum cover set. A control device for a wireless sensor network that allocates an operating period. The method of claim 8,
The processor randomly selects one of the transmitting nodes when a plurality of transmitting nodes are in an exclusive neighbor relationship to the backbone node, and the selected transmitting node transmits data to the backbone node. A control device for a wireless sensor network that allocates an operating period. The method of claim 7,
When the processor allocates the operation period of each backbone node according to the backbone scheduling algorithm, the processor collects information on the operation period of the backbone nodes that overhear data among the neighboring nodes of the transmission node, and overhears the data On the basis of the operation period information of the backbone nodes, the operation period is allocated from the backbone nodes of the lowest layer,
When the operation period of the backbone node is allocated, the operation period of each backbone node is allocated so that data collision does not occur with reference to the operation period that neighbor nodes adjacent to the corresponding backbone node overhear. The method of claim 10,
When allocating the operation period of the backbone node, the processor allocates an operation period in which a backbone node receiving data transmission from a transmission node to which the first operation period is allocated transmits data to its parent node.
When the parent node operates in an active slot that does not overhear one or more transmitting nodes to which the first operation period is assigned, the backbone node allocates a first operation period to the operation of transmitting data to the parent node,
When the parent node operates in an active slot that eavesdrops on one or more transmitting nodes to which the first operation period is assigned, the wireless sensor assigns a second operation period to the operation of transmitting data to the parent node by the backbone node. The control device of the network.

Images