WiFi-based electronic tag Locating Algorithm

WiFi-based electronic tag Locating Algorithm

With the rapid development of wireless communication, the combination of indoor positioning wireless networks and RFID technology has become increasingly concerned. People's demand for items and personnel positions is getting stronger and stronger. Outdoor positioning, such as well-known GPS positioning, has already satisfied many people, but once they enter the room, due to the blocking of buildings and the multi-path effect, the effect of GPS in indoor positioning is greatly reduced, so indoor positioning research has become the focus of subsequent research. There is a large scope of positioning personnel and items in a company. The traditional tag positioning distance is flawed, which limits its wide application. Therefore, RFID technology and wireless networks are proposed to expand their positioning scope.

Wireless WiFi in a free GHz band delivers high data transmission speeds. Therefore, select the location tag Based on Wi-Fi network communication. The Wi-Fi network has the following advantages: Wi-Fi is in the 2.4 GHz operating band and is in the free band, so no extra fee is required for users. Wi-Fi transmission distance can reach 100 m, it can cover the entire building. The Wi-Fi transmission rate is very high and can reach 54 Mbps.

Positioning accuracy is influenced not only by the selection of positioning technology, but also by the selection of positioning algorithms. Common indoor positioning algorithms are classified into two types: ranging-based and distance-independent algorithms. Generally, the distance or angle between nodes is used to calculate the location of an unknown Node Based on the ranging technology. In practical use, the following common algorithms are used: the algorithm based on the received signal strength indication (ECC) AOA and TOA. Distance-independent algorithms include the centroid method, APIT algorithm, and convex planning algorithm. These algorithms use the proximity between nodes for locating.

Generally, the algorithm based on the ranging technology is higher than that without ranging. This paper uses RFID technology based on wireless networks, and proposes an algorithm based on this to implement a location system with a small error range.

System hardware structure

Radio Frequency Identification (RFID) is commonly known as electronic tags. RFID is a non-contact automatic identification technology. It uses RF signals to automatically identify the target object and obtain relevant data. Without manual intervention, it can work in various harsh environments. RFID technology can recognize high-speed moving objects and recognize multiple tags at the same time for quick and convenient operations. RFID is a simple wireless system with only two basic devices. It is used to control, detect, and track objects. The system consists of an inquiry device (or reader) and many Responder (or tag.

The positioning system hardware includes: reader, electronic tag, and Wireless WiFi module.

A reader is a device used to read/write tag information.

Electronic Tags are classified into active and passive tags. The active technology electronic tag has a battery inside, and its life is generally longer than that of the passive one. Send messages to the outside through the set frequency band before the battery replacement. The active technology electronic tags used in this paper have a long life.

The wireless wi-fi module is mainly used for communication between electronic tags, readers, and AP (used to receive the emission signals of tags.

RFID positioning can be used for accurate positioning of warehouse management, company personnel, items, and hospital patients. However, due to the limited distance, Wireless WiFi technology and RFID technology are combined to further improve the positioning range and accuracy. The system hardware structure 1 is shown in.

System software and Positioning Algorithm

1) algorithm based on signal strength

Traditional signal transmission is susceptible to refraction, reflection, diffraction, and diffraction. the received signal strength is the superposition of signals transmitted through various channels. So sometimes the signal strength increases and sometimes decreases. After a large number of practices, it is found that the strength of the received signal is subject to the log-normal distribution. The distance between nodes is estimated by signal attenuation in transmission. The signal field strength at the undetermined position is obtained based on the channel model:

Formula: n is the path loss index, which is related to the surrounding environment; X Σ is a normal random variable with the standard deviation of Σ; d0 is the reference distance, usually 1 m in the indoor environment; PL (d0) is the signal strength of the reference position.

Assume that there are n AP and m Reference tags, then the AP receives the strong metric P = (AP1, AP2 ,..., APn), the intensity vector of the t reference tag collected is St = (St1, St2 ,..., The Euclidean distance between the tag to be determined and the reference tag St is:

The LANDMARC algorithm is represented by the signal strength algorithm. This algorithm mainly finds the reference tag closest to the location of the tag to be determined by comparing different Et. When a tag to be tested is determined by K adjacent reference tags, we call it the "K-Nearest Neighbor Algorithm". The coordinates of the tag to be determined are (x, y ):

The Wi and (xi, yi) are the weights and coordinates of the I-th neighbor reference tag. Based on experience:

The larger the weight, the smaller the E value.

Although the LANDMARC embedding method can handle complex environments, multi-path effects may occur in some closed environments, resulting in low positioning accuracy. Some researchers have proposed improvements to the landm arc algorithm: they have added the signed values of tags obtained from different readers to a set, then, find the tag with the highest frequency in the set as the tag with the closest distance, and then use the empirical formula to find the coordinates of the tag to be tested. In this way, more accurate precision can be obtained.

2) three-sided Positioning Algorithm

Three-way Positioning Method: circle the three APs at known locations, respectively, and circle the distance from the nearest reference tag to the tag to be tested as the radius. The intersection of the obtained three circles is D. Triangle algorithm 2.

Set the position node D (x, y). The coordinates of A, B, and C are (x1, y1), (x2, y2), (x3, y3 ). The distance between them and D is d1, d2, and d3. Then, the position of D can be solved by any two of the following equations.

However, in practical application, due to the existence of measurement errors, it is difficult to place the three circles at one point. This is often the case, which leads to the absence of solutions to the equation and the inability to locate the target to be tested.

3) algorithms used in this article

In this solution, the positioning algorithm we use is based on the LANDMARC algorithm (the LANDMARC algorithm), and the three-way positioning algorithm is used behind the LANDMARC algorithm to make it more accurate.

Before the experiment, an electronic tag (reference tag) is assigned to each of the three rooms in the corridor and three rooms of A Company building, and an AP is placed at the southeast and North corners of the building. Do a Good Job of wireless communication between the upper computer and the lower computer (the connection between the software program server and the client ).

During the experiment, when the tag to be tested enters the AP (4) range, it receives the signal strength from the tag to be tested and passes it into the host computer. At the same time, it also receives the signal strength of each reference tag in each AP and transmits it to the upper computer.

The positioning algorithm builds the field strength of the tag to be tested on four APS (AP1, AP2, AP3, and tags) into a field intensity vector, and the reference tag is also built into a field intensity vector. The LANDMARC algorithm compares the Euclidean distance between the field strength vector of the tag to be tested and the reference tag field intensity vector to find the reference tags with the minimum Euclidean distance, the location of the three reference labels is also known (the reference labels are recorded at the beginning of the experiment ). For the three reference points, the radius is no longer determined based on the signal strength, but three are centered around the reference point, use the distance between the closest reference labels (to determine the placement of a reference label every few meters) as the radius to make three circles, so that the possibility of the intersection of the three circles increases.

Since it is difficult for three circles to intersect at the same point, there are three relationships between the three circles:

◆ The three circles intersect and the three circles have common areas;

◆ The three networks overlap, but there is no public area;

◆ The three circles do not overlap.

The relationship is as follows:

① When three circles have a public area, there must be three intersections in the public area. If three intersections are used as triangles, the coordinates of the tags to be tested are the coordinates of the triangle.

② When there is no public area in the intersection, there must be two public areas. Take the midpoint of the two intersections in the intersection area of the two circles, and then make a triangle with the three midpoint. The inner is the inner coordinate of the label to be tested.

③ Discard when the three circles do not intersect. Accept the last three reference tags of the next group. If the intersection is not found three times, use the positions of the three reference labels as triangles, inside it is the location of the tag to be tested.

The advantage of this algorithm is that the positioning accuracy of the original LANDMARC algorithm is further improved by performing triangle positioning. In addition, further triangular positioning based on the distance between tags can reduce additional calculations and reduce repeated measurements due to changes in the field strength of the reference tags.

Conclusion

This article mainly discusses the signal strength algorithm and the third-side non-ranging algorithm, and further improves the LANDMARC algorithm. According to the relevant experiment results, the algorithm can reach an error of about 1.5 m in positioning accuracy. This solution is suitable for a wide range of applications.