CN102325345B - A container logistics tracking and positioning method based on tag sensor network - Google Patents

Abstract

The traditional RFID (Radio Frequency Identification)-based logistic management method cannot satisfy the requirements on arranging and positioning of containers, interior environment monitoring of the containers, cargo safety and the like. Specific to the problems of low RFID data transmission amount and limited RFID communication capacity, the invention provides a container logistic tracking and positioning method based on a tag sensor network, and the method can be used for expanding an RFID communication range and simultaneously learning interior environment information and location information of the containers in real time; by connecting RFID as well as a WSM (Watchguard System Manager) and a Wi-Fi (Wireless Fidelity) network, a novel W-KNN (Weighted k-Nearest Neighbor) positioning algorithm based on neighbor weight attributes is provided to realize logistic tracking and positioning; meanwhile, a specific technical scheme and step flows are designed and are obviously different from the traditional container positioning and tracking method. The container logistic tracking and positioning method has beneficial effects in the aspects of communication range, data transmission, accurate positioning and the like.

Description

A container logistics tracking and positioning method based on tag sensor network

technical field

The present invention is a new method for container logistics tracking and positioning in RFSN (Radio Frequency Sensor Network, label sensor network), which belongs to Internet of Things, RFID (Radio Frequency Identification, radio frequency identification), WSN (Wireless Sensor Network, wireless sensor Internet) technology intersection.

Background technique

Digital logistics management needs to realize the rapid identification and positioning of containers and the intelligent monitoring of container information. It can monitor the storage environment and location information of goods in real time, thereby saving operating costs and losses.

At present, RFID technology is mainly used to enter the information of container goods, generally including the type, quantity, loading and unloading time of goods, as well as information such as loading location and unloading destination; for container placement on ships, container positioning and internal environment monitoring of containers , Cargo safety and other aspects cannot meet the requirements. The WSN system can realize node environment monitoring and positioning, but it cannot fully reflect system information. Combine the two to achieve complementary advantages, thereby greatly improving the efficiency of logistics tracking and positioning.

Combining WSN and RFID technologies to form a wireless network with stronger functions requires designing an appropriate network architecture. Due to the inconsistency of WSN and RFID communication frequency bands, different communication protocols, the amount of transmitted data is small, the tag transmits internal information while the sensor collects external information, and the energy and communication capabilities of tag nodes are limited. It is necessary to combine WSN and RFID networks Plans and techniques for analysis and research.

In this method, the intelligent gateway can effectively shield the difficulty of direct physical communication. The communication inside the RFID network and the WSN communication are independent of each other. Each network can run its own network protocol. When the nodes in the two networks need to communicate, Then communicate with a node through the gateway node.

The traditional triangulation method has dimensional redundancy, and if the distance is directly calculated, it will cause a large deviation. This dimensional redundancy mainly arises from the relative positional relationship between anchor nodes. When the spatial distance between two anchor nodes is relatively close (as shown in Figure 5(a)), the observed values of RSSI (Received Signal Strength Indication, Received Signal Strength Indication) of the two anchor nodes will have a high positive correlation . Considering the extreme case where two anchor nodes are completely coincident, the RSSI values of the two anchor nodes measured from any position will be equal, that is, their RSSI values are completely positively correlated. When the two anchor nodes are on both sides of the test area (as shown in Figure 5(b)), the observed values of the RSSI of the two anchor nodes will have a high negative correlation. If two anchor nodes are at infinity in opposite directions, the observed values of the RSSI of the two anchor nodes will be completely negatively correlated. Therefore, in response to this problem, the present invention proposes a novel W-KNN (Weighted k-Nearest Neighbor, attribute weighted k-nearest neighbor positioning algorithm) positioning algorithm based on the tag sensor network.

Contents of the invention

Technical problem: the purpose of the present invention is mainly aimed at container logistics management, proposes a kind of tracking and location method based on tag sensor network (RFSN, Radio Frequency Sensor Networks), wherein location algorithm adopts W-KNN algorithm.

Technical solution: First, several definitions are given:

Wi-Fi-based RFSN network: (Wi-Fi-RFSN, Wireless Fidelity Radio FrequencySensor Networks), using the Wi-Fi network to achieve mutual communication between all RFID readers, and then through the Wi-Fi-WSN gateway to realize RFID Communication between the network and the WSN network.

Wi-Fi-WSN gateway: It includes a wireless network card, a WSN wireless transceiver, and a data processing unit, which are used to connect the Wi-Fi network and the WSN network, and realize the communication between the two networks.

Wi-Fi-RFID reader: add a wireless network card function module to the traditional RFID reader, and realize mutual communication through a wireless access point (AP), hereinafter referred to as RFID reader or reader device.

Monitoring Center: Responsible for real-time monitoring and collection of all APs, sensor nodes, RFID readers, RFID tags and other information on the ship, and report to the master control center.

W-KNN (Weighted k-Nearest Neighbor) positioning algorithm: attribute-weighted k-nearest neighbor positioning algorithm, which is a classification algorithm for container positioning. Calculate the correlation coefficient for all anchor nodes in the test area, calculate the weight of each anchor node (distributed in the entire ship space) by the correlation coefficient, and finally use the sum of weighted distance squares to represent the Euclidean distance between two nodes.

Positioning node: fixed on the ship, the spatial position is known, with the positioning node as the reference point, after calculating the relative distance between it and the container sensing node according to the W-KNN positioning algorithm, the spatial coordinate value of the sensing node is obtained, and the positioning is completed.

WSN clustering method: WSN base station nodes traverse the neighbor list, select the neighbor node with high energy value and short response time as the first layer cluster head node, and establish the first layer clustering; the first layer cluster head traverses the neighbor list, selects the energy value Some neighbor nodes with high energy value and short response time are relay nodes; the relay node traverses the neighbor list, selects the neighbor nodes with high energy value and short response time as the second layer cluster head nodes, and establishes the second layer clustering; repeat The above steps complete the establishment of the whole WSN network clustering process.

Wi-Fi-RFSN data packet format: as shown in Figure 6, it includes two parts: data packet header and data content; among them, the data packet header is 96 bits long (12 bytes), including 8-bit message type, 8-bit total Length, 16-bit area ID, 16-bit message destination, 16-bit message source, 16-bit destination address, 16-bit source address, the data packets mentioned below follow this format.

Message type: There are 9 different data packet messages in this method, each message name and its corresponding field value, as shown in Table 1.

Total length: The total length refers to the length of the sum of the header and data content, and the unit is byte. The total length field is 8 bits, so the maximum length of the packet is 256 bytes.

Area ID: Since each area contains only one positioning node, the area ID can be represented by the area positioning node ID.

Message destination: The transmission of data packets needs to be forwarded through multi-hop routes, so the message destination needs to be indicated, that is, the object that needs to receive the data packet in the end.

Source of the message: indicates the object that originally sent the data packet.

Destination address: A and B communicate directly without going through intermediate routing and forwarding. When A sends a data packet to B, the address of B becomes the destination address.

Source address: A's address becomes the source address.

(Note: The destination address and source address can be the ID of the monitoring center, the ID of the gateway, the ID of the wireless AP, the ID of the RFID reader, the ID of the RFID tag, the ID of the WSN base station, the ID of the anchor node, and the ID of the positioning node , the ID of the container sensor node, etc.)

The invention utilizes the RFSN network to realize functions such as reading label information in the container, collecting sensor node information, and positioning the container. Before that, RFSN needs to be initialized.

Before network deployment, each ship has a monitoring center, a Wi-Fi-WSN gateway, multiple APs, multiple Wi-Fi-RFID readers, and some neatly arranged containers. Each container is affixed with an ID-marked RFID tag and installed with an ID-marked sensor node, both of which have the same identity. Each label stores an identity (ID number), identifies the type of goods in the container, the place of departure, the place of shipment, batch and other information. Each node stores an identity (ID number), energy value, cluster head node (the ID number of a node, indicating which cluster the node belongs to), node type identification (identified as cluster head, relay, common three One), the area to which it belongs. Each node also contains a neighbor list, which records information such as the ID number, energy value, cluster head number and response time of the neighbor node.

After the above information is configured, the network deployment starts.

Method flow

The complete process of the broadcast tracking and positioning method will be specifically described as follows:

The radio frequency identification (RFID) network and the wireless sensor network (WSN) are connected through a wireless compatible certification Wi-Fi network, and the attribute weighted k-nearest neighbor positioning algorithm W-KNN is used to realize the positioning of the container. The method flow can be described as follows:

1. Network initialization

Step 1) arrange and install monitoring center, Wi-Fi-WSN gateway, all containers, RFID tags, sensor nodes, Wi-Fi wireless access point AP, Wi-Fi-RFID reader-writer and anchor node;

Step 2) At this point, all APs form a Wi-Fi wireless network, and all sensor nodes form a WSN network according to the Leach algorithm. The base station stores a functional node identity ID table, storing the ID numbers of anchor nodes and positioning nodes, as Prepare for collecting location information below;

Step 3) Divide the entire test space into m*n test areas according to the distribution of positioning nodes, each test area has a unique positioning node, so the available area positioning node ID represents the area ID, and the spatial coordinates of the positioning nodes are known, where m Indicates the number of layers, the height of each layer is the height of a container, and n represents the number of test areas equally divided by each layer;

Step 4) All m*n positioning nodes broadcast the "area positioning" data packet, wherein the message type is "area positioning" and the total length is 12 bytes, indicating that the data packet only contains the header, the data content part is empty, and the area The ID is empty, the destination of the message is empty, the source of the message is empty, and the destination address is (0110101000000000) ₂ = (j0) ₁₆ , which means that the objects receiving the data packet are all container sensor nodes in this area, and the source address is the location Node ID;

Step 5) After the container sensing node receives the "area positioning" data packet, it saves the source address, that is, the positioning node ID of the area, indicating that the container sensing node belongs to the area where the positioning node is located;

Step 6) For each test area, a data training is carried out, that is, in this area, the positioning node is used as a test sample, and the sample is responsible for collecting the RSSI values of all anchor nodes in this test area to generate "RSSI value Response" data packet, in which the message type is "RSSI value response", the total length is greater than 12 bytes, the area ID is the ID of the positioning node in the area, the destination of the message is the ID of the monitoring center, the source of the message is the ID of the positioning node, and the data content is Multi-group data in the format of "anchor node ID, RSSI value", and transmit all RSSI values to the cluster head node in the cluster, and the cluster head node uploads it to the base station node step by step, and finally the base station node passes Wi-Fi - WSN gateway transmits to the monitoring center;

为区域a内锚节点i与锚节点j的相关系数，i＝1，2，...，k_a，j＝1，2，...，k_a，a＝1，2，...，m*n，且 $r_{\mathrm{ij}}^a=\frac{\mathrm{Cov}({\mathrm{rssi}}_i^a,{\mathrm{rssi}}_j^a)}{\sqrt{D\left({\mathrm{rssi}}_i^a\right)*D\left({\mathrm{rssi}}_j^a\right)}}=\frac{E({\mathrm{rssi}}_i^a*{\mathrm{rssi}}_j^a)-E\left({\mathrm{rssi}}_i^a\right)*E\left({\mathrm{rssi}}_j^a\right)}{\sqrt{E\left[{\left({\mathrm{rssi}}_j^a\right)}^2\right]-{\left[E\left({\mathrm{rssi}}_j^a\right)\right]}^2}*\sqrt{E[{\left({\mathrm{rssi}}_i^a\right)}^2-{\left[E\left({\mathrm{rssi}}_j^a\right)\right]}^2}},$ 当i＝j时，

Step 7) The monitoring center receives multiple sets of "RSSI value response" data packets M _rssi in m*n areas, and uses the W-KNN positioning algorithm to calculate the correlation coefficient matrix R in units of each area, and there are m*n in total , it may be assumed that there are k anchor nodes in the area a, a=1, 2,..., m*n, then the corresponding correlation coefficient matrix is

is the correlation coefficient between anchor node i and anchor node j in area a, i=1, 2,..., k _a , j=1, 2,..., k _a , a=1, 2,... , m*n, and $r_{\mathrm{ij}}^a=\frac{\mathrm{Cov}({\mathrm{rssi}}_i^a,{\mathrm{rssi}}_j^a)}{\sqrt{\mathrm{D.}\left({\mathrm{rssi}}_i^a\right)*\mathrm{D.}\left({\mathrm{rssi}}_j^a\right)}}=\frac{\mathrm{E.}({\mathrm{rssi}}_i^a*{\mathrm{rssi}}_j^a)-\mathrm{E.}\left({\mathrm{rssi}}_i^a\right)*\mathrm{E.}\left({\mathrm{rssi}}_j^a\right)}{\sqrt{\mathrm{E.}\left[{\left({\mathrm{rssi}}_j^a\right)}^2\right]-{\left[\mathrm{E.}\left({\mathrm{rssi}}_j^a\right)\right]}^2}*\sqrt{\mathrm{E.}[{\left({\mathrm{rssi}}_i^a\right)}^2-{\left[\mathrm{E.}\left({\mathrm{rssi}}_j^a\right)\right]}^2}},$ When i=j,

和

分别表示区域a内测试样本对锚节点i与锚节点j的RSSI的观测值，E(X)和D(X)分别表示随机变量X的数学期望和方差，Cov(X，Y)表示随机变量X和Y的协方差；若两个锚节点的可检测范围没有交集即它们从来没有被同时检测到过，或其中至少一个锚节点的RSSI的观测值的方差为0，则定义该锚节点对的相关系数为0；然后，计算区域a内锚节点i的权值为

其中，

表示

的转置，即

并且将处理后的数据存储到后台数据库的“区域定位信息数据表”中；in,

and

Represent the observed values of the RSSI of the test sample in area a to anchor node i and anchor node j, E(X) and D(X) respectively represent the mathematical expectation and variance of the random variable X, Cov(X, Y) represent the random variable The covariance of X and Y; if the detectable range of two anchor nodes has no intersection, that is, they have never been detected at the same time, or the variance of the RSSI observation value of at least one anchor node is 0, then define the anchor node pair The correlation coefficient of is 0; then, calculate the weight of anchor node i in area a as

in,

express

the transposition of

And store the processed data in the "area positioning information data table" of the background database;

2. Information collection process

Step 8) The monitoring center broadcasts the "information collection" data packet, wherein the message type is "information collection" and the total length is 12 bytes, indicating that the data packet only contains the header, the data content part is empty, the area ID is empty, and the message destination If it is empty, the source of the message is the ID of the monitoring center, and the destination address is all 0, indicating that the objects receiving the data packet are all receiving devices in the entire Wi-Fi-RFSN network, and the source address is the ID of the monitoring center;

Step 9) After the Wi-Fi-WSN gateway receives the "information collection" data packet, the data packet is converted into WSN data packet format and Wi-Fi data packet format, and sent to the wireless AP of WSN base station and Wi-Fi respectively;

2.1 Collect container sensor node information

Step 10) After the WSN base station receives the "information collection" message from the gateway, it broadcasts the "sensing information collection" data packet in the entire WSN network, wherein the message type is "sensing information collection", and the total length is 12 bytes. Indicates that the data packet only contains the header, the data content part is empty, and the destination and destination addresses of the message are both (0110101000000000) ₂ = (j0) ₁₆ , indicating that the objects receiving the data packet are all container sensor nodes in the entire WSN network, and the message The source is the monitoring center ID, and the source address is the base station node ID;

Step 11) After all container sensing nodes receive the "sensing information collection" data packet, they collect the internal environment information of the container and generate a "sensing collection response" data packet, wherein the message type is "sensing collection response", The total length is greater than 12 bytes, the area ID is the location node ID of the area where the sensor node of the container is located, the destination of the message is the ID of the monitoring center, the source of the message is the ID of the sensor node of the container, and the destination address is the cluster where the sensor node of the container is located The head node ID, the source address is the container sensor node ID, and the data content is the internal environment information of the container collected by the container sensor node;

Step 12) The container sensor node sends the "sensing collection response" data packet to the cluster head node in the cluster, and the cluster head node transmits the data packet to the upper cluster head node through the relay node until it is sent to the base station node ;

Step 13) The base station node sends the "sensing collection response" data packet to the Wi-Fi-WSN gateway. At this time, the destination address is the gateway ID, the source address is the WSN base station ID, and the rest of the messages remain unchanged;

Step 14) The Wi-Fi-WSN gateway sends the "sensing collection response" data packet to the monitoring center. At this time, the destination address is the monitoring center ID, the source address is the gateway ID, and the remaining messages remain unchanged;

Step 15) the monitoring center stores the received data packet in the "sensor sensing information data table" of the background database;

2.2 Collect container RFID tag information

Step 16) After the Wi-Fi wireless AP receives the "information collection" message from the gateway, it broadcasts the "label information collection" data packet in the entire network, that is, within the Wi-Fi signal coverage area, wherein the message type is "label information collection" , the total length is 12 bytes, which means that the data packet only contains the header, the data content part is empty, the area ID is empty, and the destination of the message is (0110001000000000) ₂ = (b0) ₁₆ , which means that the object that finally receives the data packet is Wi- For all RFID tags in the entire Fi network, the source of the message is the ID of the monitoring center, and the destination address is (0111100100000000) ₂ = (y0) ₁₆ , indicating that the object receiving the data packet is all readers in the entire Wi-Fi network, and the source address is The ID of the wireless AP that sent the message;

Step 17) After the Wi-Fi-RFID reader receives the "tag information collection" message from the wireless AP, it sends electromagnetic waves. After the RFID tag receives the electromagnetic wave, the coil generates an induced current, and the information built in the tag is emitted through the electromagnetic wave;

Step 18) After the reader receives the electromagnetic wave of the tag, it converts it into a data packet suitable for communication in the Wi-Fi network, which is a "tag collection response" data packet, wherein the message type is "tag collection response", and the total length More than 12 bytes, the area ID is empty, the destination of the message is the ID of the monitoring center, the source of the message is the container RFID tag ID, the destination address is the wireless AP that can communicate directly with the reader, and the source address is the Wi-Fi that reads the tag information - The ID of the RFID reader, the content of the message is the built-in information of the RFID tag of the container;

Step 19) Communicate between wireless APs, and finally send the "label collection response" data packet to the Wi-Fi-WSN gateway. At this time, the destination address is the gateway ID, and the source address is the ID of the AP closest to the gateway. The rest of the messages are not Change;

Step 20) The Wi-Fi-WSN gateway sends the "label collection response" data packet to the monitoring center. At this time, the destination address is the monitoring center ID, the source address is the gateway ID, and the rest of the messages remain unchanged;

Step 21) the monitoring center stores the received data packet in the "RFID tag information data table" of the background database;

2.3 Collect container location information

Step 22) After the WSN base station receives the "information collection" message from the gateway, it broadcasts the "location information collection" data packet in the entire WSN network, wherein the message type is "location information collection", and the total length is 12 bytes, indicating that the data The packet only contains the header, the data content part is empty, the area ID is empty, and the message destination and destination address are both (0110110100000000) ₂ = (m0) ₁₆ , indicating that the objects receiving the data packet are all anchor nodes in the entire WSN network, The source of the message is the monitoring center ID, and the source address is the base station node ID;

Step 23) After the anchor node receives the "location information collection" data packet, it broadcasts the data packet;

Step 24) The positioning node and the container sensor node send a "position acquisition response" data packet to the cluster head node in the cluster where the message type is "position acquisition response", the total length is greater than 12 bytes, and the area ID is the container The ID of the positioning node in the area where the sensor node is located, the destination of the message is the ID of the monitoring center, the source and source address of the message are both the ID of the positioning node or the container sensor node, and the destination address is the cluster head of the cluster where the positioning node or container sensor node is located Node ID, the data content is the received signal strength of all anchor nodes in the area, that is, the RSSI value;

Step 25) the cluster head node transmits the data packet to the cluster head node of the upper layer through the relay node until it is sent to the base station node;

Step 26) The base station node sends the "position acquisition response" data packet to the Wi-Fi-WSN gateway. At this time, the destination address is the gateway ID, the source address is the WSN base station ID, and the rest of the messages remain unchanged;

Step 27) The Wi-Fi-WSN gateway sends the "position acquisition response" data packet to the monitoring center. At this time, the destination address is the monitoring center ID, the source address is the gateway ID, and the rest of the messages remain unchanged;

Step 28) The monitoring center stores the received data packet in the "sensing node/positioning node position information data table" of the background database;

3. Information processing process

3.1 Calculation of location information

表示区域a内第i个锚节点的权值，

分别表示区域a内定位节点e、节点f对第i个锚节点的RSSI观测值；已知区域a定位节点的坐标为(x，y，z)，x表示横坐标，y表示纵坐标，z表示竖坐标，则集装箱传感节点f的空间坐标为(x，y+d(e，f)，z)；监控中心将计算处理后的的数据存储到后台数据库的“传感器坐标信息数据表”中；Step 29) The monitoring center finds the "sensing node/positioning node position information data table" from the background database, and divides the data in the table into m*n groups according to the area ID, and each group uses the W-KNN positioning algorithm to calculate the positioning The distance d(e, f) between the node and all container nodes in the area, $d(e,f)=\sqrt{{\mathrm{\Σ}}_{i=1}^{k_a}{w}_i^a{({\mathrm{rssi}}_{e,i}^a-{\mathrm{rssi}}_{f,i}^a)}^2},$ Among them, e represents the positioning node in area a, f represents the container sensor node in area a and f=1, 2, ..., l _a ; l _a represents the number of container nodes in area a, and k _a represents the number of container nodes in area a number of anchor nodes,

Indicates the weight of the i-th anchor node in area a,

Respectively represent the RSSI observation values of the positioning node e and node f in the area a to the i-th anchor node; the coordinates of the positioning node in the known area a are (x, y, z), x represents the abscissa, y represents the ordinate, z Indicates the vertical coordinate, then the spatial coordinate of the container sensor node f is (x, y+d(e, f), z); the monitoring center stores the calculated and processed data in the "sensor coordinate information data table" of the background database middle;

3.2 Synthesis of all information

Step 30) The monitoring center finds "sensor sensing information data table", "RFID tag information data table" and "sensor coordinate information data table" from the background database, according to "container sensor node ID and container RFID tag ID one by one The rule of "corresponding" combines the data in the three tables into a "container data" information, and stores it in the "container information data table" of the background database, where the container ID is the ID or RFID tag of the sensor node in the container The ID of the container message contains three parts of information, namely sensor sensing information, RFID tag information and sensor location coordinate information;

4. Information sending process

Step 31) The monitoring center sends the "container data" message to the communication satellite through the GSM module; the communication satellite forwards it to the receiving tower on the ground; Track location messages.

Beneficial effects: the present invention proposes a container logistics tracking and positioning method based on a tag sensor network, which has the following advantages:

(1) By using the tag sensor network to track and locate the logistics, the problem of the short communication distance of the RFID reader is overcome, the communication range is expanded, and combined with the application of the sensor node, the content of the collected information is improved, and the container can be more comprehensively understood environmental conditions;

(2) By using the W-KNN positioning algorithm combined with the sensor nodes, the positioning of the container on the ship can be realized, and the anchor node weights are allocated according to the correlation, which improves the positioning accuracy to a certain extent.

Description of drawings

Fig. 1 Schematic diagram of the structure of the container logistics tracking and positioning method based on RFSN,

Figure 2 RFSN instance layout structure diagram,

Fig. 3 RFSN network topology diagram,

Fig. 4 Container positioning top view,

Figure 5 The relative position relationship diagram of anchor nodes,

Figure 6 Wi-Fi-RFSN packet format diagram.

Detailed ways

The following will take the accompanying drawings as an example to further describe the technical solution and method flow of the present invention in detail through a specific example.

1. Network initialization

Step 1) Arrange and install the monitoring center, Wi-Fi-WSN gateway, all containers (with RFID tags and sensor nodes installed), AP, Wi-Fi-RFID readers, anchor nodes and positioning nodes, Figure 2 shows the layout and installation Post-RFSN instance layout structure diagram;

Step 2) At this point, all APs form a Wi-Fi wireless network, and all sensor nodes form a WSN network according to the Leach algorithm, as shown in the network topology of Figure 3; nodes d1 and d2 in Figure 2 are positioning nodes, and node m1 , m2, and m3 are anchor nodes, then the base station stores a functional node ID table that records the positioning nodes and anchor nodes in the format of [location node, d1, d2], [anchor node, m1, m2, m3], which is collected as follows Prepare location information;

Step 3) In Fig. 2, the space of the whole ship is divided into 2*1 test areas according to the distribution of positioning nodes, wherein, containers 1, 2, and 3 are in area 1, and containers 4, 5, and 6 are in area 2; Each test area has a positioning node, the positioning node of area 1 is d1, and the positioning node of area 2 is d2;

Step 4) All positioning nodes (2*1) in Fig. 2 broadcast "area positioning" data packets, the format is as shown in Figure 6, taking the positioning node d1 in area 1 as an example, it broadcasts data packets M _qd =[ "Area location", 12, null, null, null, j0, d1, null], where "Area location" is the message type, 12 indicates that the total length is 12 bytes, and three nulls indicate that the area ID is empty and the message The destination is empty, the source of the message is empty, j0 is the destination address (indicating that the object receiving the data packet is all the container sensor nodes in the area), d1 is the source address, and the last null indicates that the data content part is empty;

Step 5) After the container sensing node in Fig. 2 receives the "area positioning" data packet M _qd , taking the container sensing nodes j1, j2, j3 in area 1 as an example, they save the source address d1, indicating that the container belongs to Locate the area where node d1 is located;

Step 6) For each test area, conduct a data training. Take area 1 in Figure 2 as an example, take the positioning node d1 as the test sample, and d1 is responsible for collecting all anchor nodes (m1, m2, m3) in area 1. RSSI value, generate "RSSI value response" packet M _rssi = ["RSSI value response", >12, d1, jc, d1, c2, d1, (mi, 5)], wherein "RSSI value response" is the message type ,>12 means the total length is greater than 12 bytes, and the data content is not empty. The three d1 are the area ID, the source of the message, and the source address, jc is the destination of the message, c2 is the destination address, (mi, 5) is the content of the message, and mi The values of m1, m2, and m3 (representing the anchor node ID of the area), 5 is the measured RSSI value; as shown in Figure 3, it may be assumed that node d1 belongs to cluster 2, and d1 transmits all regional positioning response messages M _rssi to The cluster head node c2 in the cluster, the cluster head node c2 is transmitted to the relay node z1, z1 is transmitted to the cluster head node c1 of the upper layer, and c1 is transmitted to the base station node, and finally the base station node passes through the Wi-Fi-WSN gateway sent to the monitoring center;

为锚节点mi与锚节点mj的相关系数，且 $r_{\mathrm{ij}}^1=\frac{\mathrm{Cov}({\mathrm{rssi}}_i^1,{\mathrm{rssi}}_j^1)}{\sqrt{D\left({\mathrm{rssi}}_i^1\right)*D\left({\mathrm{rssi}}_j^1\right)}}=\frac{E({\mathrm{rssi}}_i^1*{\mathrm{rssi}}_j^{\text{1}})-E\left({\mathrm{rssi}}_i\right)*E\left({\mathrm{rssi}}_j^1\right)}{\sqrt{E\left[{\left({\mathrm{rssi}}_j^1\right)}^2\right]-{\left[E\left({\mathrm{rssi}}_j^1\right)\right]}^2}*\sqrt{E[{\left({\mathrm{rssi}}_i^1\right)}^2-{\left[E\left({\mathrm{rssi}}_j^1\right)\right]}^2}},$ (i＝1，2，3、j＝1，2，3)，其中，

和分别表示测试样本d1对锚节点mi与锚节点mj的RSSI的观测值，E(X)和D(X)分别表示随机变量X的数学期望和方差，Cov(X，Y)表示随机变量X和Y的协方差；不妨假设计算得到区域1的相关系数矩阵 $R^1=\left(\begin{array}{ccc}1& \frac{1}{2}& \frac{1}{3}\\ \frac{1}{2}& 1& 1\\ \frac{1}{3}& \frac{1}{4}& 1\end{array}\right);$ 然后，计算区域1内所有锚节点的权值，节点m1的权值为 $w_1^1=\frac{1}{R_1^1*{\left(R_1^1\right)}^T}=\frac{1}{(1\·\frac{1}{2}\·\frac{1}{2})*{(1\·\frac{1}{2}\·\frac{1}{2})}^T}=\frac{1}{1+\frac{1}{4}+\frac{1}{9}}=\frac{36}{49},$ 同理，得到 $w_2^1=\frac{16}{21},w_3^1=\frac{144}{169};$ 并且按照

的格式，存储到后台数据库中，表示定位节点d1所在的区域内的锚节点m1的权值为

该区域内的锚节点m2、锚节点m3情况类似，其余区域的情况也类似，此处不再赘述；Step 7) The monitoring center jc in Figure 2 receives multiple sets of "RSSI value response" data packets M _rssi in 2*1 areas, and uses the W-KNN positioning algorithm to calculate the correlation coefficient matrix R (total 2*1), taking area 1 as an example, there are 3 anchor nodes in area 1, and the corresponding correlation coefficient matrix is $R^1=\left(\begin{array}{ccc}{r}_{11}^1& r_{12}^1& r_{13}^1\\ r_{\mathrm{twenty\; one}}^1& r_{\mathrm{twenty\; two}}^1& r_{\mathrm{twenty\; three}}^1\\ r_{31}^1& r_{32}^1& r_3^1\end{array}\right),$

is the correlation coefficient between anchor node mi and anchor node mj, and $r_{\mathrm{ij}}^1=\frac{\mathrm{Cov}({\mathrm{rssi}}_i^1,{\mathrm{rssi}}_j^1)}{\sqrt{\mathrm{D.}\left({\mathrm{rssi}}_i^1\right)*\mathrm{D.}\left({\mathrm{rssi}}_j^1\right)}}=\frac{\mathrm{E.}({\mathrm{rssi}}_i^1*{\mathrm{rssi}}_j^{\text{1}})-\mathrm{E.}\left({\mathrm{rssi}}_i\right)*\mathrm{E.}\left({\mathrm{rssi}}_j^1\right)}{\sqrt{\mathrm{E.}\left[{\left({\mathrm{rssi}}_j^1\right)}^2\right]-{\left[\mathrm{E.}\left({\mathrm{rssi}}_j^1\right)\right]}^2}*\sqrt{\mathrm{E.}[{\left({\mathrm{rssi}}_i^1\right)}^2-{\left[\mathrm{E.}\left({\mathrm{rssi}}_j^1\right)\right]}^2}},$ (i=1, 2, 3, j=1, 2, 3), where,

and Represent the observed values of the test sample d1 on the RSSI of the anchor node mi and the anchor node mj, E(X) and D(X) represent the mathematical expectation and variance of the random variable X, Cov(X, Y) represent the random variable X and The covariance of Y; let's assume that the correlation coefficient matrix of area 1 is calculated $R^1=\left(\begin{array}{ccc}1& \frac{1}{2}& \frac{1}{3}\\ \frac{1}{2}& 1& 1\\ \frac{1}{3}& \frac{1}{4}& 1\end{array}\right);$ Then, calculate the weights of all anchor nodes in area 1, and the weight of node m1 is $w_1^1=\frac{1}{R_1^1*{\left(R_1^1\right)}^T}=\frac{1}{(1\&Center\; Dot;\frac{1}{2}\&Center\; Dot;\frac{1}{2})*{(1\&Center\; Dot;\frac{1}{2}\·\frac{1}{2})}^T}=\frac{1}{1+\frac{1}{4}+\frac{1}{9}}=\frac{36}{49},$ In the same way, get $w_2^1=\frac{16}{\mathrm{twenty\; one}},w_3^1=\frac{144}{169};$ and follow

The format is stored in the background database, indicating that the weight of the anchor node m1 in the area where the positioning node d1 is located is

The situation of anchor nodes m2 and anchor nodes m3 in this area is similar, and the situation of other areas is also similar, so I won’t repeat them here;

2. Information collection process

Step 8) Monitoring center broadcasts "information collection" data packet M _xc =["information collection", 12, null, null, jc, 00, jc], wherein, "information collection" is a message type, and 12 represents a total length of 12 byte (indicating that the data packet only contains the header, and the data content part is empty), two nulls respectively indicate that the area ID is empty and the destination of the message is empty, two jc are the source and source address of the message respectively, 00 (hexadecimal number ) that is, the 16-bit number of the destination address is all 0 (indicating that the object receiving the data packet is all receiving devices in the entire Wi-Fi-RFSN network);

Step 9) After the Wi-Fi-WSN gateway receives the "information collection" data packet M _xc , the data packet is converted into the WSN data packet format M _xcs and the Wi-Fi data packet format M _xcf , and sent to the WSN base station and the Wi-Fi data packet format M xcf respectively. Fi's wireless AP;

2.1 Collect container sensor node information

Step 10) After the WSN base station receives the "information collection" data packet M _xcs of the gateway, it broadcasts the "sensing information collection" data packet M _cc in the entire WSN network = ["sensing information collection", 12, null, j0, jc, j0, jz], where "sensing information collection" is the message type, 12 indicates that the total length is 12 bytes (indicating that the data packet only contains the header, and the data content part is empty), null indicates that the area ID is empty, and two A j0 represents the destination and destination address of the message respectively (representing that the objects receiving the data packet are all container sensor nodes in the entire WSN network), jc represents the original source of the message, and jz represents the source address;

Step 11) After all the container sensor nodes receive the "sensing information collection" data packet M _cc , they collect the internal environment information of the container, generate the "sensing collection response" data packet M _ccx , and use the container sensor nodes in Fig. 3 Take j1 as an example, j1 generates a data packet M _ccx = ["Sensing Collection Response", >12, d1, jc, j1, c2, j1, 34°C], where "Sensing Collection Response" is the message type, >12 Indicates that the total length is greater than 12 bytes, the data content is not empty, d1 is the area ID, jc is the destination of the message, two j1 are the source and source address of the message, c2 is the destination address, and 34°C (take this as an example to illustrate the container sensor The internal environment information of the container collected by the node) is the message content;

Step 12) The container sensor node j1 sends the "sensing collection response" data packet M _ccx to the cluster head node c2 in the cluster where it is located, and c2 transmits the data packet to the upper cluster head node c1 through the relay node z1 until sending to the base station node jz;

Step 13) The base station node sends the "sensing collection response" data packet to the Wi-Fi-WSN gateway, M _ccx = ["sensing collection response", >12, d1, jc, j1, wg, jz, 34°C] , at this time the gateway wg is the destination address, jz is the source address, and the rest of the messages remain unchanged;

Step 14) The Wi-Fi-WSN gateway sends the "sensing collection response" data packet to the monitoring center, M _ccx = ["sensing collection response", >12, d1, jc, j1, jc, wg, 34°C] , at this time, the monitoring center jc is both the message destination and the destination address, the gateway wg is the source address, and the rest of the messages remain unchanged;

Step 15) The monitoring center jc stores it in the background database in the format of [j1, 34°C], indicating that the temperature inside the container where the container node j1 is located is 34°C;

2.2 Collect container RFID tag information

Step 16) After the Wi-Fi wireless AP receives the "information collection" message M _xcf of the gateway, it broadcasts the "label information collection" data packet M _bc in the entire Wi-Fi network, and takes the wireless access point a3 in Fig. 3 as Example, M _bc =["tag information collection", 12, null, b0, jc, y0, a3], wherein, "tag information collection" is the message type, and 12 represents that the total length is 12 bytes (representing that the data packet only contains The first part, the data content part is empty), null means that the area ID is empty, b0 is the destination of the message (indicating that the object that finally receives the data packet is all RFID tags in the Wi-Fi network), the monitoring center jc is the source of the message, and y0 is Destination address (indicating that the objects receiving the data packet are all readers in the Wi-Fi network), a3 is the source address:

Step 17) Take reader y1 and tag b1 in Fig. 3 as an example, after y1 receives the "tag information collection" message M _bc from the wireless AP, it sends an electromagnetic wave, and after the RFID tag b1 receives the electromagnetic wave, the coil generates an induced current, and the The information built into the tag is emitted through electromagnetic waves;

Step 18) After the reader y1 receives the electromagnetic wave of the tag b1, it converts it into a data packet suitable for communication in the Wi-Fi network, which is the "tag collection response" data packet M _bcx = ["tag collection response", >12 , null, jc, b1, a1, y1, "Place of origin=Beijing, place of destination=Nanjing"], where "Label Collection Response" is the message type, >12 means the total length is greater than 12 bytes, and the data content is not empty , null indicates that the area ID is empty, the monitoring center jc is the destination of the message, the source RFID tag b1 is the source of the message, the wireless access point (AP) a1 is the destination address, and the Wi-Fi-RFID reader that transmits Wi-Fi wireless signals y1 is the source address, "place of departure=Beijing, place of destination=Nanjing" (the built-in information of the container RFID tag) is the message content;

Step 19) Communicate between wireless APs, and finally send the "label collection response" data packet M _bcx to the Wi-Fi-WSN gateway wg, M _bcx = ["label collection response", >12, null, jc, b1, wg , a3, "place of departure=Beijing, place of destination=Nanjing"], the gateway wg is the destination address, the wireless access point a3 closest to the gateway is the source address, and the rest of the messages remain unchanged;

Step 20) The Wi-Fi-WSN gateway wg sends the "label collection response" data packet M _bcx to the monitoring center jc, M _bcx = ["label collection response", >12, null, jc, b1, jc, wg, " Place of departure=Beijing, place of destination=Nanjing”], at this moment, the monitoring center jc is both the destination and the destination address of the message, the gateway wg is the source address, and the rest of the messages remain unchanged;

Step 21) The monitoring center jc stores in the background database according to the format of [b1, "place of departure=Beijing, place of destination=Nanjing"], indicating that the place of departure of the container where the label b1 is located is Beijing, and the place of shipment is Nanjing;

2.3 Collect container location information

Step 22) After the WSN base station receives the "information collection" M _xcs message of the gateway, it broadcasts the "location information collection" data packet M _wc =["location information collection", 12, null, m0, jc, m0 in the entire WSN network. , jz], wherein, "location information collection" is the message type, 12 indicates that the total length is 12 bytes (indicating that the data packet only contains the header, and the data content part is empty), null indicates that the area ID is empty, and the two m0 are respectively The destination and destination address of the message (indicating that the objects receiving the data packet are all anchor nodes in the entire WSN network), the monitoring center jc is the source of the message, and the base station node jz is the source address;

Step 23) After the anchor node receives the "location information collection" data packet M _wc , it broadcasts the data packet;

Step 24) In Fig. 3, the positioning node d1 and the container sensor node j1 send the "position collection response" data packet M _wcx = ["position collection response", > 12, d1, jc, d1 to the cluster head node c2 of the cluster where they are located (or j1), c2, d1 (or j1), rssi _d1 (or rssi _j1 )], wherein, "Position Collection Response" is the message type, >12 means the total length is greater than 12 bytes, the data content is not empty, and d1 is The area ID, the monitoring center jc is the destination of the message, the two d1 (or j1) are the source and source address of the message respectively, the cluster head node c2 where it is located is the destination address, rssi _d1 (or rssi _j1 ) (represents the message received from the anchor node signal strength) is the data content;

Step 25) The cluster head node c2 transmits the data packet to the cluster head node c1 of the upper layer through the relay node z1 until it is sent to the base station node jz;

Step 26) The base station node jz sends the "position acquisition response" data packet M _wcx to the Wi-Fi-WSN gateway, M _wcx = ["position acquisition response", >12, d1, jc, d1 (or j1), wg, jz, rssi _d1 (or rssi _j1 )], now the gateway wg is the destination address, the WSN base station jz is the source address, and the rest of the messages remain unchanged;

Step 27) The Wi-Fi-WSN gateway wg sends the "position acquisition response" data packet M _wcx to the monitoring center jc, M _wcx = ["position acquisition response", >12, d1, jc, d1 (or j1), jc , wg, rssi _d1 (or rssi _j1 )], the monitoring center jc is the destination address, the gateway wg is the source address, and the rest of the messages remain unchanged;

Step 28) The monitoring center jc is stored in the background database according to the format of [d1 (or j1), d1, rssi _d1 (or rssi _j1 )], indicating that in the area where the positioning node d1 is located, the positioning node d1 (or container sensing Node j1) received anchor node signal strength (ie RSSI value);

3. Information processing process

3.1 Calculation of location information

分别表示区域1内定位节点e、集装箱传感节点f对第i个锚节点的RSSI观测值；不妨设区域1内的定位节点d1对3个锚节点的RSSI值分别为2，3，4，集装箱节点j1对3个锚节点的RSSI值分别为3，3，4，则d1与j1之间的距离为 $d(d1,j1)=\sqrt{{\mathrm{\Σ}}_{i=1}^3{w}_i^1{({\mathrm{rssi}}_{e,i}^1-{\mathrm{rssi}}_{f,i}^1)}^2}=$ $\sqrt{\frac{36}{49}*{(3-2)}^2+\frac{16}{21}*{(3-3)}^2+\frac{144}{169}*{(4-4)}^2}=\frac{6}{7};$ 已知区域1中定位节点d1的坐标为(5，4，3)，x表示横坐标，y表示纵坐标，z表示竖坐标，则集装箱传感节点f的空间坐标为

监控中心jc按照的格式存储到后台数据库中，表示集装箱传感节点j1的空间坐标为

区域1内的集装箱节点j2、集装箱节点j3情况类似，其余区域的情况也类似，此处不再赘述；Step 29) The monitoring center finds the table shown in (d) in Table 2 from the background database, and divides the data in the table into 2*1 groups according to the area ID, and each group uses the W-KNN positioning algorithm to calculate The distance d(e, f) between the positioning node and all container nodes in the area, as shown in Figure 4, taking area 1 as an example, $d(e,f)=\sqrt{{\mathrm{\Σ}}_{i=1}^3{w}_i^1{\left({\mathrm{rssi}}_{e,i}^1-{\mathrm{rssi}}_{f,i}^1\right)}^2},$ Where e represents the positioning node in area 1, f represents the container sensor node in area 1 and f=1, 2, 3, Indicates the weight of the i-th anchor node in area 1,

respectively represent the RSSI observation values of the positioning node e and the container sensor node f in the area 1 to the i-th anchor node; let the RSSI values of the positioning node d1 in the area 1 to the three anchor nodes be 2, 3, 4, The RSSI values of the container node j1 to the three anchor nodes are 3, 3, and 4 respectively, so the distance between d1 and j1 is $d(d1,j1)=\sqrt{{\mathrm{\Σ}}_{i=1}^3{w}_i^1{({\mathrm{rssi}}_{e,i}^1-{\mathrm{rssi}}_{f,i}^1)}^2}=$ $\sqrt{\frac{36}{49}*{(3-2)}^2+\frac{16}{\mathrm{twenty\; one}}*{(3-3)}^2+\frac{144}{169}*{(4-4)}^2}=\frac{6}{7};$ Given that the coordinates of the positioning node d1 in area 1 are (5, 4, 3), x represents the abscissa, y represents the ordinate, and z represents the vertical coordinate, then the spatial coordinates of the container sensor node f are

The monitoring center jc follows The format is stored in the background database, indicating that the spatial coordinates of the container sensor node j1 are

The situation of container node j2 and container node j3 in area 1 is similar, and the situation of other areas is also similar, so I won’t repeat them here;

3.2 Synthesis of all information

其中，j1/b1(集装箱内传感节点/RFID标签ID)表示集装箱ID，是集装箱消息内容，包含三部分信息，即传感器传感信息、RFID标签信息和传感器位置坐标信息；Step 30) The monitoring center jc finds the tables of the form (b), (c), and (e) in Table 2 from the backstage database, and will The data in the three tables is merged into one "container data" message

Among them, j1/b1 (sensing node/RFID tag ID in the container) represents the container ID, It is the container message content, including three parts of information, namely sensor sensing information, RFID tag information and sensor location coordinate information;

4. Information sending process

Step 31) the monitoring center jc sends "container data" _Mgd to the communication satellite through the GSM module; the communication satellite forwards it to the receiving tower on the ground; This real-time tracking positioning message.

The following table 1 is a table of message types and corresponding field values:

serial number message name message field value 1 Regional positioning 00000001 2 RSSI value response 00000010 3 Information Collection 00000011 4 Sensor information collection 00000100 5 Sensing collection response 00000101 6 Label information collection 00000110 7 Label Collection Response 00000111 8 Location Information Collection 00001000 9 Location collection response 00001001

The following table 2 is the database storage data format table:

serial number Regional positioning node ID Anchor Node ID Anchor node weight 1 2

(a) Regional positioning information data table

serial number container sensor node ID Sensing message content 1 2

(b) Sensor Sensing Information Data Sheet

serial number container RFID tag ID Label message content 1 2

(c) RFID tag information data sheet

(d) Sensor node/positioning node position information data table

serial number container sensor node ID Spatial coordinates 1 2

(e) Sensor coordinate information data table

serial number Container ID container message content 1 2

(f) Container Information Data Sheet

Claims (1)

1. A container logistics tracking and location method based on a label sensor network, characterized in that a radio frequency identification RFID network and a wireless sensor network WSN are connected through a wireless compatible authentication Wi-Fi network, and attribute weighted k nearest neighbors are used to locate Algorithm W-KNN realizes container positioning, and its method flow can be described as follows: 1. Network initialization Step 1) Arrange and install the monitoring center, Wi-Fi-WSN gateway, all containers, RFID tags, sensor nodes, Wi-Fi wireless access point AP, Wi-Fi-RFID reader and anchor node; Step 2) At this point, all APs form a Wi-Fi wireless network, and all sensor nodes form a WSN network according to the Leach algorithm. The base station stores a functional node ID table, storing the ID numbers of anchor nodes and positioning nodes, as Prepare for collecting location information below; Step 3) Divide the entire test space into m*n test areas according to the distribution of positioning nodes, each test area has a unique positioning node, so the available area positioning node ID represents the area ID, and the spatial coordinates of the positioning nodes are known, where m Indicates the number of layers, the height of each layer is the height of a container, and n represents the number of test areas equally divided by each layer; Step 4) All m*n positioning nodes broadcast "area positioning" data packets, in which the message type is "area positioning" and the total length is 12 bytes, indicating that the data packet only contains the header, the data content part is empty, and the area The ID is empty, the destination of the message is empty, the source of the message is empty, and the destination address is (0110101000000000) ₂ = (j0) ₁₆ , which means that the objects receiving the data packet are all container sensor nodes in this area, and the source address is the location Node ID; Step 5) After the container sensor node receives the "area positioning" data packet, it saves the source address, that is, the location node ID of the area, indicating that the container sensor node belongs to the area where the location node is located; Step 6) For each test area, a data training is performed, that is, the positioning node is used as a test sample in this area, and the sample is responsible for collecting the received signal strength indication RSSI value of all anchor nodes in this test area to generate "RSSI value Response" data packet, in which the message type is "RSSI value response", the total length is greater than 12 bytes, the area ID is the ID of the positioning node in the area, the destination of the message is the ID of the monitoring center, the source of the message is the ID of the positioning node, and the data content is Multi-set data in the format of "anchor node ID, RSSI value", and transmit all RSSI values to the cluster head node in the cluster, and the cluster head node uploads it to the base station node step by step, and finally the base station node passes Wi-Fi- The WSN gateway sends it to the monitoring center; Step 7) The monitoring center receives multiple sets of "RSSI value response" data packets M _rssi in m*n areas, and uses the W-KNN positioning algorithm to calculate the correlation coefficient matrix R in units of each area, and there are m*n in total , it may be assumed that there are k _a anchor nodes in area a, a=1,2,...,m*n, then the corresponding correlation coefficient matrix is

is the correlation coefficient between anchor node i and anchor node j in area a, i=1,2,...,k _a , j=1,2,...,k _a , a=1,2,...,m*n, and $r_{\mathrm{ij}}^a=\frac{\mathrm{Cov}({\mathrm{rssi}}_i^a,{\mathrm{rssi}}_j^a)}{\sqrt{\mathrm{D.}\left({\mathrm{rssi}}_i^a\right)*\mathrm{D.}\left({\mathrm{rssi}}_j^a\right)}}=\frac{\mathrm{E.}({\mathrm{rssi}}_i^a*{\mathrm{rssi}}_j^a)-\mathrm{E.}\left({\mathrm{rssi}}_i^a\right)*\mathrm{E.}\left({\mathrm{rssi}}_j^a\right)}{\sqrt{\mathrm{E.}\left[{\left({\mathrm{rssi}}_i^a\right)}^2\right]-{\left[\mathrm{E.}\left({\mathrm{rssi}}_i^a\right)\right]}^2}*\sqrt{\mathrm{E.}\left[{\left({\mathrm{rssi}}_j^a\right)}^2\right]-{\left[\mathrm{E.}\left({\mathrm{rssi}}_j^a\right)\right]}^2}},$ When i=j,

in,

and

respectively represent the observed values of the RSSI of the test sample in area a to anchor node i and anchor node j, E(X) and D(X) respectively represent the mathematical expectation and variance of the random variable X, and Cov(X,Y) represent the random variable The covariance of X and Y; if the detectable range of two anchor nodes has no intersection, that is, they have never been detected at the same time, or the variance of the RSSI observation value of at least one anchor node is 0, then define the anchor node pair The correlation coefficient of is 0; then, calculate the weight of anchor node i in area a as $w_i^a=\frac{1}{R_i^a*{\left(R_i^a\right)}^T},$ in, $R_i^a=(r_{i1}^a,r_{i2}^a,...,r_{{\mathrm{ik}}_a}^a),$

express

the transposition of

And store the processed data in the "area positioning information data table" of the background database; 2. Information collection process Step 8) The monitoring center broadcasts the "information collection" data packet, in which the message type is "information collection" and the total length is 12 bytes, indicating that the data packet only contains the header, the data content part is empty, the area ID is empty, and the destination of the message If it is empty, the source of the message is the ID of the monitoring center, and the destination address is all 0, indicating that the objects receiving the data packet are all receiving devices in the entire Wi-Fi-RFSN network, and the source address is the ID of the monitoring center; Step 9) After the Wi-Fi-WSN gateway receives the "information collection" data packet, it converts the data packet into WSN data packet format and Wi-Fi data packet format, and sends them to the WSN base station and Wi-Fi wireless AP respectively; 2.1 Collect container sensor node information Step 10) After receiving the "information collection" message from the gateway, the WSN base station broadcasts the "sensing information collection" data packet in the entire WSN network. The message type is "sensing information collection" and the total length is 12 bytes. Indicates that the data packet only contains the header, the data content part is empty, and the destination and destination addresses of the message are both (0110101000000000) ₂ =(j0) ₁₆ , indicating that the objects receiving the data packet are all container sensor nodes in the entire WSN network, and the message The source is the monitoring center ID, and the source address is the base station node ID; Step 11) After receiving the "sensing information collection" data packet, all container sensor nodes collect the internal environment information of the container and generate a "sensing collection response" data packet, wherein the message type is "sensing collection response", The total length is greater than 12 bytes, the area ID is the location node ID of the area where the sensor node of the container is located, the destination of the message is the ID of the monitoring center, the source of the message is the ID of the sensor node of the container, and the destination address is the cluster where the sensor node of the container is located The head node ID, the source address is the container sensor node ID, and the data content is the internal environment information of the container collected by the container sensor node; Step 12) The container sensor node sends the "sensing collection response" data packet to the cluster head node in the cluster, and the cluster head node transmits the data packet to the upper cluster head node through the relay node until it is sent to the base station node ; Step 13) The base station node sends the "sensing collection response" data packet to the Wi-Fi-WSN gateway. At this time, the destination address is the gateway ID, the source address is the WSN base station ID, and the rest of the messages remain unchanged; Step 14) The Wi-Fi-WSN gateway sends the "sensing collection response" data packet to the monitoring center. At this time, the destination address is the monitoring center ID, the source address is the gateway ID, and the rest of the messages remain unchanged; Step 15) The monitoring center stores the received data packets in the "sensor sensing information data table" of the background database; 2.2 Collect container RFID tag information Step 16) After receiving the "information collection" message from the gateway, the Wi-Fi wireless AP broadcasts the "label information collection" data packet in the entire network, that is, within the Wi-Fi signal coverage area, where the message type is "label information collection" , the total length is 12 bytes, indicating that the data packet only contains the header, the data content part is empty, the area ID is empty, and the destination of the message is (0110001000000000) ₂ = (b0) ₁₆ , indicating that the object that finally receives the data packet is Wi- For all RFID tags in the entire Fi network, the source of the message is the ID of the monitoring center, and the destination address is (0111100100000000) ₂ =(y0) ₁₆ , which means that the object receiving the data packet is all readers in the entire Wi-Fi network, and the source address is The ID of the wireless AP that sent the message; Step 17) After the Wi-Fi-RFID reader receives the "tag information collection" message from the wireless AP, it sends electromagnetic waves. After the RFID tag receives the electromagnetic wave, the coil generates an induced current, and the information built in the tag is emitted through the electromagnetic wave; Step 18) After the reader receives the electromagnetic wave of the tag, it converts it into a data packet suitable for communication in the Wi-Fi network, which is the "tag collection response" data packet, where the message type is "tag collection response", and the total length More than 12 bytes, the area ID is empty, the destination of the message is the ID of the monitoring center, the source of the message is the container RFID tag ID, the destination address is the wireless AP that can communicate directly with the reader, and the source address is the Wi-Fi that reads the tag information - The ID of the RFID reader, the content of the message is the built-in information of the RFID tag of the container; Step 19) Communicate between wireless APs, and finally send the "label collection response" data packet to the Wi-Fi-WSN gateway. At this time, the destination address is the gateway ID, the source address is the ID of the AP closest to the gateway, and the rest of the messages are not Change; Step 20) The Wi-Fi-WSN gateway sends the "label collection response" data packet to the monitoring center. At this time, the destination address is the monitoring center ID, the source address is the gateway ID, and the rest of the messages remain unchanged; Step 21) The monitoring center stores the received data packets in the "RFID tag information data table" of the background database; 2.3 Collect container location information Step 22) After receiving the "information collection" message from the gateway, the WSN base station broadcasts the "location information collection" data packet in the entire WSN network. The message type is "location information collection" and the total length is 12 bytes, indicating the data The packet only contains the header, the data content part is empty, the area ID is empty, and the message destination and destination address are both (0110110100000000) ₂ = (m0) ₁₆ , which means that the objects receiving the data packet are all anchor nodes in the entire WSN network, The source of the message is the monitoring center ID, and the source address is the base station node ID; Step 23) After the anchor node receives the "location information collection" data packet, it broadcasts the data packet; Step 24) The positioning node and the container sensor node send a "position acquisition response" data packet to the cluster head node in the cluster where the message type is "position acquisition response", the total length is greater than 12 bytes, and the area ID is the container The ID of the positioning node in the area where the sensor node is located, the destination of the message is the ID of the monitoring center, the source and source address of the message are both the ID of the positioning node or the container sensor node, and the destination address is the cluster head of the cluster where the positioning node or container sensor node is located Node ID, the data content is the received signal strength of all anchor nodes in the area, that is, the RSSI value; Step 25) The cluster head node transmits the data packet to the upper cluster head node through the relay node until it is sent to the base station node; Step 26) The base station node sends the "location collection response" data packet to the Wi-Fi-WSN gateway. At this time, the destination address is the gateway ID, the source address is the WSN base station ID, and the rest of the messages remain unchanged; Step 27) The Wi-Fi-WSN gateway sends the "location collection response" data packet to the monitoring center. At this time, the destination address is the monitoring center ID, the source address is the gateway ID, and the rest of the messages remain unchanged; Step 28) The monitoring center stores the received data packets in the "sensing node/positioning node position information data table" of the background database; 3. Information processing process 3.1 Calculation of location information Step 29) The monitoring center finds the "sensing node/positioning node position information data table" from the background database, and divides the data in the table into m*n groups according to the area ID, and each group uses the W-KNN positioning algorithm to calculate the positioning The distance d(e, f) between the node and all container nodes in the area,

Where e represents the positioning node in area a, f represents the container sensor node in area a and f=1, 2,..., l _a ; l _a represents the number of container nodes in area a, k _a represents the anchor node in area a number,

Indicates the weight of the i-th anchor node in area a, Respectively represent the RSSI observation values of the positioning node e and node f in the area a to the i-th anchor node; the coordinates of the positioning node in the known area a are (x, y, z), x represents the abscissa, y represents the ordinate, and z Indicates the vertical coordinate, then the spatial coordinate of the container sensor node f is (x, y+d(e, f), z); the monitoring center stores the calculated and processed data in the "sensor coordinate information data table" of the background database middle; 3.2 Synthesis of all information Step 30) The monitoring center finds the "sensor sensing information data table", "RFID tag information data table", and "sensor coordinate information data table" from the background database, and according to the "container sensor node ID and container RFID tag ID one by one The rule of "corresponding" combines the data in the three tables into a "container data" information, and stores it in the "container information data table" of the background database, where the container ID is the ID or RFID tag of the sensor node in the container The ID of the container message contains three parts of information, namely sensor sensing information, RFID tag information and sensor location coordinate information; 4. Information sending process Step 31) The monitoring center sends the "container data" message to the communication satellite through the GSM module; the communication satellite then forwards it to the receiving tower on the ground; Track location messages.

Images