KR102089113B1 - Congestion control method at large scale crop management system based on wireless multimedia sensor network - Google Patents

Abstract

The present invention proposes an improved application protocol that can efficiently transmit image information as well as various scalar sensing data in a large-scale cultivation system based on ICT, and defines interactions with network protocols to ensure the quality of service of image information. .
According to the present invention, image transmission efficiency is improved and PSNR performance is improved.

Description

Service quality (QoS) guaranteed image transmission method in large-scale cultivation system based on ICT {CONGESTION CONTROL METHOD AT LARGE SCALE CROP MANAGEMENT SYSTEM BASED ON WIRELESS MULTIMEDIA SENSOR NETWORK}

The present invention relates to a quality of service (QoS) guaranteed image transmission protocol in a large-scale cultivation system based on ICT.

Research on systems that can increase production efficiency and reduce crop damage caused by natural disasters or pests by applying ICT (Information and Communications Technology) technology to large-scale greenhouses or open-field crop growth environments has been actively conducted at home and abroad.

Cultivation in large scale greenhouse environment on / humidity, as well as simple numerical data, such as CO ₂ is present, the image data received by CCTV and the camera comes. Unlike simple scalar data, image data has a relatively large data size and delay-sensitive characteristics. ICT-based large-scale cultivation system basically consists of various sensors that collect information related to the cultivation environment and growth. Recently, the deployment of multimedia sensors capable of capturing image and image information for the diagnosis and production of pests and pests has increased.

Conventional cultivation management system collects environmental information in the form of scalar for information related to the cultivation environment such as temperature and humidity, and a simple monitoring service is provided based on this, but recently, a variety of sophisticated and intelligent control of the cultivation environment is provided. Sensors and actuators were deployed. As the arrangement of various sensors and the like increases, a change occurs in the traffic pattern transmitted from the ICT-based cultivation system, and in particular, a mechanism for efficiently transmitting multimedia information such as images must be added.

In a wireless sensor network (WSN) based cultivation management system, there is a difficulty in efficient image transmission. The cause is due to the limitations of the image processing structure and communication method, which are not suitable for large data processing and transmission. Since the WSN operates based on low-power wireless communication, the radio link state is relatively unstable, so the probability of transmission error is high, so the packet error rate is high. In addition, when a specific event occurs, traffic rapidly increases, and congestion occurs, so that a lot of data is discarded. In the case of important data, retransmission is required for packets that are discarded due to transmission errors or congestion. Therefore, the energy efficiency due to retransmission is lowered, and there arises a problem of using resources inefficiently.

Therefore, there is a need for a communication structure and protocol for energy-efficient and quality of service for WMSN-based image information transmission.

Republic of Korea Patent Publication 2010-0021057 (2010.02.24)

Therefore, the present invention was devised to solve the problems of the prior art as described above, and the general purpose of the present invention is ICT-based large-scale cultivation that can substantially compensate for various problems caused by limitations and disadvantages in the prior art. The system is to provide a quality of service (QoS) guaranteed image transmission protocol.

Another specific object of the present invention is to efficiently transmit various scalar sensing data as well as image information (multimedia data) by an application layer protocol in a large-scale cultivation system based on ICT, and image information (multimedia) by interaction with a network layer protocol. Data) to provide a service quality (QoS) guaranteed image transmission method in a large-scale cultivation system based on ICT that can guarantee service quality.

To this end, the present invention proposes an improved application layer protocol capable of efficiently transmitting image information (multimedia data) as well as various scalar sensing data in a large-scale cultivation system based on ICT, and guarantees service quality (delay and error rate) of multimedia data. In order to do so, it defines the interaction with the network layer protocol.

According to the results of the performance analysis through the simulation method, it was found that the mechanism proposed in the present invention improves the efficiency of image transmission and improves PSNR performance.

1 is a flowchart illustrating a quality of service (QoS) guaranteed image transmission protocol in an ICT-based large-scale cultivation system according to an embodiment of the present invention.
2 is a diagram showing a network configuration of a simulation according to an embodiment of the present invention.
3 is a graph showing an average packet delay time result.
4 is a graph showing packet loss rate results.
5 is a graph showing PSNR comparison results.

Efficient transmission of multimedia information such as images is difficult in WSN due to the following two reasons.

First, since the WSN is operated based on low-power wireless communication, since the radio link state is relatively unstable, the probability of transmission error is high, so the packet error rate is high. The image can be accurately reproduced by the sequential combination of multiple packets, and the high packet error rate makes it difficult for the receiver to properly reproduce the image.

Second, WSN is relatively prone to congestion. Because the density of nodes is very high, when a specific event occurs, multiple sensor nodes detect the event and transmit traffic at the same time. Therefore, the traffic is rapidly increased instantaneously, and congestion occurs, and a situation in which a large number of packets are discarded frequently occurs.

The requirements that a communication protocol must support for efficient image transmission in WSN are as follows.

○ The level of reliability and quality of service required for image reproduction must be guaranteed.

○ In order to provide reliability of image information, a retransmission mechanism is required for errors or lost packets.

○ Because WSN has resource limitations such as memory, retransmission is performed at the end-to-end application protocol level rather than hop-by-hop retransmission.

○ At the network layer, the QoS guarantee mechanism, which is differentiated according to the importance of data, must be supported.

○ When a network congestion occurs, it should include a function that preferentially delivers information of high priority such as image packets.

○ The retransmission protocol must have an asymmetrical structure that performs complex functions in the sink node without resource constraints and performs minimal functions in the sensor node.

○ In consideration of the WSN situation in which computing resources such as CPU performance are insufficient and battery supply and replacement are difficult, the communication protocol must be high in energy efficiency and the protocol implementation should be optimized.

Therefore, additional mechanisms are needed for communication protocols such as the network layer for image transmission in WSN.

The present invention analyzes the requirements of a communication protocol for reliable image transmission, defines the functions of the network layer and application layer based on this, and proposes an improved protocol. The simulation method is used to analyze the performance of the proposed protocol.

2. QoS  Proposed image transmission protocol

2.1 Function definition of each layer protocol

(1) Application layer protocol function

The application layer protocol supports a reliable information transmission function between application processes of two end points performing communication by using a service provided at the network layer. The role of the application protocol in terms of transmitting image information of WSN is to transmit and receive all image information without error in order between the image sensor capturing the image information and the application process of the sink node collecting the image information. WSN transfers information over the air between sensor nodes, and since the transmission rate is relatively low, it is impossible to transmit the entire image information at once. Therefore, the transmitting side divides one image information into several packets, and each segmented packet is independently transmitted to the sink node. In addition to the segmented image information, control information such as sequence number information and error detection information indicating the order of each packet is additionally attached to each packet. The sink node, which is the final destination of the image information, uses the sequence number information of the received packet to grasp information about the packet that has been lost in a certain period, and requests retransmission of the lost or damaged packets to the transmitter.

In addition to reliable information transmission, the application layer of WSN classifies sensing data according to importance and determines the transmission policy according to the importance of data. Since the density of WSN nodes is very high, there is a high possibility that adjacent nodes sense a specific event in duplicate, and the sensor node is equipped with various sensors according to its role, so the types and characteristics of the sensed data are very diverse. Considering the resource limit of the sensor node and the low data rate, it is efficient to transmit the sensed data differentially according to importance and characteristics. The classification of sensing data is primarily classified into scalar data and multimedia data.

Scalar data means environmental information such as temperature and humidity, and is composed of simple numerical values, so the data capacity is relatively small. , Multimedia data is image, sound, and video information. The data capacity is very large, and a series of values are correlated to represent information. Also, at the application layer level, it is classified into a connected application and a connectionless application according to whether a logical connection (session) is established at the application layer level of the source node and the sink node. In multimedia information, it is necessary to maintain a logical connection (session) because data of a certain period is highly correlated. In contrast, Scala data is a connectionless service that does not require connection maintenance.

It is classified into the following four types in consideration of the sensed data characteristics.

○ Energy efficiency: best effort

○ Reliability type: It is not sensitive to delay, but requires reliable data transmission

○ Real-time guarantee: Reliability is not very important, but data must be transmitted within the deadline.

○ Reliability-real-time guarantee type: High reliability as well as important data that must be transmitted within the deadline

(2) Network layer protocol function

The main role of the network layer is to determine the transmission path of the packet from the source sensor node to the destination node (sink node), and perform the function of transferring the packet to the sink node through multi-hop wireless communication through the corresponding path. . As a core function of the network layer of routing, an optimal route to a destination is selected in consideration of geographical characteristics, link status, buffer occupancy, and the like. In the initial WSN, the most important and considered factor in routing and packet delivery was to minimize energy consumption. Recently, as the WSN function is expanded, a mechanism capable of guaranteeing service quality has emerged. In particular, since there is a high possibility of a congestion situation, there is a growing demand for a routing technique that prevents congestion in advance and a function capable of preferentially delivering important packets when congestion occurs.

Therefore, in the network layer, the packets delivered from the application layer are considered in consideration of importance and network status, thereby guaranteeing a differentiated quality of service, and transmitting packets to the sink node while supporting overall network energy efficiency.

At the network layer, it can be classified into the need for retransmission, real-time transmission, and best transmission according to the importance of the packet, and the importance of the packet is determined at the application layer, and based on this information, it is reflected in the packet importance classification and routing at the network layer. . The packet forwarding function is a function of forwarding a packet to a destination at a relay node, and a differentiated delivery policy may be applied according to the importance of the packet. If congestion occurs at the relay node, packets with low importance may inevitably be discarded in the middle.

(3) Interface definition between layers

The interface for packet transmission and reception between the application layer and the network layer is divided into two types according to a connected service and a connectionless service.

The interface for supporting a connectionless service is as follows.

○ CL_DATA_REQ

-Request packet transmission to network layer without state management

-Parameters: packet content, data importance, traffic type

○ CL_DATA_IND

-Notification of packet reception to application layer without state management

-Parameters: packet content, data importance, traffic type

The interface for supporting the connected service is as follows.

○ CO_DATA_REQ

-Maintain logical connection between two application layers and request packet transmission from network layer

-Parameters: packet content, connection identifier, data importance, traffic type

○ CO_DATA_IND

-Maintain a logical connection between the two application layers, and notify the application layer of packet reception

-Parameters: packet content, connection identifier, data importance, traffic type

○ CO_DATA_RJT

-Requests packet transmission from the application layer of the sink node to retransmit the lost packet to the network layer

-Parameters: sensor node number, connection identifier, retransmission packet sequence number

○ CO_DATA_RTX

-As a primitive transmitted from the network layer of the sensor node to the application layer, the sync node notifies the reception of a packet requesting retransmission

-Parameters: sensor node number, connection identifier, retransmission packet sequence number

2.2 Application protocol for image transmission

In the case of important data, retransmission is required for packets that are discarded due to transmission errors or congestion. The need for image information transmission in WMSN is increasing compared to the existing WSN. WMSN generates multimedia information such as images and sounds with relatively large data sizes by sensor nodes. More energy is consumed to efficiently process and transmit it. Multimedia information can be provided only when the quality of service such as delay and error rate is guaranteed and transmitted. Therefore, an energy efficient multimedia information processing function and a communication protocol are required to maximize network life while ensuring service quality. The communication protocol for image transmission requires close interaction between the application layer and the network layer considering the resource limitation of the sensor node, and the protocol structure should be optimized through the cross layer design model.

In order to minimize the capacity of the image information, it is possible to improve transmission efficiency and energy efficiency by extracting the updated object from the background image and transmitting only the extracted update object information. In WMSN applications, the camera mot has a fixed view frame. In order to detect moving objects, a background subtraction technique is used. That is, as a method of detecting an object from a difference between a current frame and a background image, the background image means a stationary camera view without a moving object, updates periodically (periodically), and uses a running Gausian average technique. .

(1) Control message definition

○ image packet (sensor-> sink)

-Purpose: A message that transmits real image data

-Parameters: Packet ID, Image data information, CRC information

○ camera setup (sink-> sensor)

-Purpose: To specify the camera setting information of the sensor node

-Parameters: parameters related to camera setting information

○ image query (sink-> sensor)

-Purpose: A message requesting summary information about an image

-Parameters: Packet ID, Image data information, CRC information

○ image summary (sensor-> sink)

-Purpose: A message that transmits summary information about image characteristics, such as image size, in response to an image query.

-Parameters: summary information such as image size

○ ACK (sensor <-> sink)

-Purpose: Affirmative response to normal receipt of message

-Parameters: Packet ID, etc.

○ NACK (sensor <-> sink)

-Purpose: Incorrect response to abnormal message reception, such as a message error, retransmission request, etc.

-Parameters: Packet ID, etc.

○ Start-of-TX (sink-> sensor)

-Purpose: A message requesting the start of image data transfer

-Parameters: Packet size and transmission policy ID

○ End-of-TX (sink-> sensor)

-Purpose: To end the image data transfer procedure

-Parameters: Packet size and transmission policy ID

(2) Message delivery flow

The first problem to be solved for image transmission is the sensitivity to packet errors. In particular, the loss of a packet corresponding to an important part of compression causes a significant problem in image reproduction in the sink. Therefore, a reliable transport protocol at the application layer is required. It is essential to send and receive all image information without error in order. The image information is divided into several packets, and a number (ID) is assigned to each packet, and additional packet control information and error detection information are added. Each packet is transmitted in order. After that, it includes procedures such as acknowledgment and retransmission.

In the conventional invention, only one sensor node can transmit image data during a specific session, but in the present invention, multiple sensor nodes can simultaneously transmit image information at a specific time.

The sensor node that acquires the image information and determines that it is necessary to transmit to the sink sends an image summary message to the sink node. Upon receiving the image summary message, the sensor node grasps the image summary information and the current network status, and when it is determined that image information reception is necessary, responds through the image request message. The sensor node receiving the image request message is divided into appropriate size packets and sequentially transmits image information to the sink node. The sink node receiving the image information transmits retransmission information to the sensor node for an error or out-of-order packet. At this time, if necessary, an acknowledgment can be sent to a certain number of packets at the sink node.

When the transmission of the segmented image packet is completed, the sink node transmits an image transmission completion message to complete the image transmission session.

2.3 Network protocol for image transmission

At the network layer, the quality level is marked on the packet based on the importance of the packet determined by the application layer. This part is based on the packet marking algorithm [~]. The quality level is classified into three levels: green, yellow, and red [~]. Each quality level is as follows.

○ Green Packet-means a packet that has the highest importance and should be guaranteed with high reliability and real time. Set the path with link quality as the top priority.

○ Yellow Packet-It has an intermediate level of importance and considers link quality and buffer occupancy at an appropriate level to establish a route.

○ Red Packet-This is the quality level marked on the packet with the lowest importance. Even if the link quality is low, it can be transmitted and may be discarded depending on the situation when congestion occurs.

In the case of multimedia data such as images, videos, and sounds, green data may be marked, and scalar data such as temperature, humidity, and illuminance may be marked in red. In the classifier of each sensor node, the quality level is marked in the packet priority field according to the importance of the corresponding packet, and a differentiated routing policy is reflected according to the marked packet to set the path to the sink node.

3. Performance analysis of image transmission protocol

3.1 Simulation Configuration

2 shows the network topology of the simulation. The structure of 10x10 square array is composed of 100 nodes, and the distance between nodes is set to 1m. The types of nodes are classified into sink node, source node, relay node, and background traffic node. The sink node is a destination node of all packets, collects and analyzes data sensed by each node, and requests retransmission when there is a problem with the packet. The source node is a node equipped with an image sensor and periodically transmits image data in packets of a certain size. Set the node farthest from the sink node. The relay node transmits data to the sink node by setting the differential path of data received from the neighbor node. The background traffic node is a node for adjusting the state of the network, and controls the congestion of the network while adjusting the packet generation cycle. In order to transmit all nodes in the simulation to the sink node, it is possible to transmit only the upper and the left adjacent.

3.2 Simulation Environment

Table 1 shows the parameters used in the simulation. The simulation was performed in the Visual Studio 2010 environment, and the simulation program was performed using C ++. The sink node is set to (0, 0) and the source node is set to (10, 10). Background traffic nodes for controlling the state of the network are nodes except the source node at the right and bottom sides, assuming the network is rectangular. The remaining nodes in the network area act as relay nodes. The transmission range is set to 1M. Between each node, a link cost value indicating link quality exists. The link cost value was set to a random value in the range (between 1.0 and 2.0 in 0.01 units), and there is a difference in transmission time according to the link cost value. When the link cost is 1.0, the transmission time is 0.96 ms, and when it is 2.0, the transmission time is 1.92 ms. The initial energy values of the nodes are set to 1000000 (1J). The energy value is 185uj for transmission, 83uj for reception, and 15uj for waiting for collision.

Variables and setting values used in simulation parameter Setting value Network size (0, 0) ~ (10, 10) Sink node (0, 0) Source node (99, 99) Background traffic node (90, 90) ~ (97, 97)
(9, 9) ~ (79, 79) Number of nodes 100 Transmission range 1M Packet size 16 bytes, 256 bytes Initial node energy 1J Transmission energy consumption 0.021Mj / Bit Energy consumption 0.014Mj / Bit Atmospheric energy consumption 64μJ Simulation time 100MS

3.3 Analysis of simulation results

The simulation compares the minimum hop routing protocol with the average packet delay, packet error rate, and data throughput of the proposed mechanism. We experimented by changing the packet generation cycle of the background traffic node from 100ms to 35ms. When the generation cycle is 100ms, it shows a smooth network state and when it is 35ms, it shows a very congested state. The source node was fixed at 100 ms. As a retransmission method, hop-by-hop retransmission and end-to-end retransmission can be set. Packets transmitted from the source node can be set in two cases: 16 bytes and 256 bytes.

3.3.1 Average Packet Delay

Timely transmission of multimedia data is important as well as the data itself. Therefore, it is inevitable to be sensitive to the delay time it takes to transmit. The delay time of the packet was determined from the time the data packet was generated to the time received at the destination sink node, and the average packet delay was defined as the average of the delay times of all packets received at the sink node. 3 shows the average packet delay time observed after the simulation is completed in the min hop and proposed mechanism. Experimental results show that the proposed protocol has a lower overall latency than the min hop routing protocol. In the proposed protocol, packet latency is reduced by 36% on average, and QoS QoS guarantee is verified.

3.3.2 Packet Error Rate

4 is a result of comparing the proposed mechanism and the min hop routing protocol for the packet loss rate. When the image is transmitted using the min hop routing protocol, packets start to be lost when the network status is congested by about 70%, and gradually increase, and when congestion reaches 85%, more than half of the images are lost. have. On the other hand, when the image is transmitted by applying the proposed mechanism, the packet loss rate is reduced by up to 93% until the network network status reaches 80%. The min hop routing protocol is very vulnerable to congestion due to the repeated shortest path selection of all packets. However, the proposed mechanism showed an effect of increasing the limit of network congestion on packet loss caused by congestion due to distributed packet routing. Through this, it was verified that the proposed mechanism was improved in the packet loss rate evaluation.

3.3.3 Data Throughput

Data throughput is an important evaluation criterion for the purpose of image transfer of the cultivation management system. The original image assuming that the RGB image of the crop was sensed by the source node and the image received by the sink node were compared using the PSNR formula. 5 shows the experimental results. The PSNR value is infinite if there is no difference from the original, but the resulting graph cannot be expressed infinity, so the maximum value is set to 50 dB for convenience. Looking at the graph, the PSNR of the min hop routing protocol gradually decreases when the network congestion exceeds 60% and decreases below 20dB. On the other hand, the proposed mechanism does not differ from the original video up to 75% network congestion, but decreases from 85% congestion to 20% or less. Through simulation, it was confirmed that the network congestion limit for perfect data processing was improved by 15% and the proposed protocol was suitable for image transmission.

Embodiments of the present invention include computer readable media including program instructions for performing various computer-implemented operations. Computer-readable media may include program instructions, local data files, local data structures, or the like alone or in combination. The media may be specially designed and constructed for the present invention, or may be known and usable to those skilled in computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical recording media such as CD-ROMs, DVDs, magnetic-optical media such as floptical discs, and ROM, RAM, flash memory, etc. Hardware devices specifically configured to store and execute the same program instructions are included. Examples of program instructions include high-level language code that can be executed by a computer using an interpreter, etc., as well as machine language codes produced by a compiler.

On the other hand, in the detailed description of the present invention and the accompanying drawings, specific embodiments have been described, but the present invention is not limited to the disclosed embodiments and the technical spirit of the present invention can be understood by those skilled in the art to which the present invention pertains. Various substitutions, modifications and changes are possible without departing from the scope.

Claims (1)

(i) classifying a communication protocol for quality of service guarantee image transmission in an ICT-based large-scale cultivation system into an application layer protocol and a network layer protocol;
(ii) classifying a plurality of sensing data to be transmitted from the source sensor node at the transmitting side by the application layer protocol into scalar data having a relatively small data capacity and multimedia data having a relatively large capacity:
(iii) determining a transmission path of a packet from the source sensor node to the sink node by selecting an optimal path to a sink node in consideration of geographic characteristics, link status, and buffer occupancy by the network layer protocol;
(iv) The multimedia data is divided into multiple packets at the transmitting side by the application layer protocol, and each of the divided packets is connected to the network layer protocol in step (iii) by interaction with the network layer protocol. Each packet is transmitted to the sink node independently by a hop-by-hop retransmission method or an end-to-end retransmission method along the transmission path of the packet determined by each packet. Allowing control information including information or error detection information to be attached; And
(v) By using the sequence number information of the packet received from the sink node, which is the final destination of multimedia data, by the application layer protocol, information on the packet lost in a certain period of time is identified to transmit the lost or damaged packets. It includes the step of requesting a retransmission,
Step (iv) includes detecting, extracting and transmitting the updated object from the background image based on the difference between the background image and the current frame included in the multimedia data, and transmitting the detected object.
Step (v) transmits a packet requesting retransmission for the lost packet by sending CO_DATA_RJT having a sensor node number, a connection identifier, and a retransmission packet sequence number as parameters in the application layer of the sink node to the transmitting side. Requesting the network layer; And a primitive transmitted from the network layer of the source sensor node to the application layer by transmitting CO_DATA_RTX having a sensor node number, a connection identifier, and a retransmission packet sequence number as parameters by the transmitting side. A method of transmitting a quality of service (QoS) -guaranteed image in a large scale cultivation system based on ICT, comprising the step of notifying the reception of a requested packet.

Images