WO2023089362A1

WO2023089362A1 - System and method of positioning inside buildings

Info

Publication number: WO2023089362A1
Application number: PCT/IB2021/060720
Authority: WO
Inventors: Przemysław Adamkiewicz
Original assignee: Centrum Badawczo Rozwojowe Technologii Informatycznych Spółka Z Ograniczoną Odpowiedzialnością; Styła, Michał
Priority date: 2021-11-18
Filing date: 2021-11-18
Publication date: 2023-05-25

Abstract

The subject of the invention is a system for determining the position inside objects. It is characterized by the fact that it consists of radio probes (1, 2, 3,..., n) working in Bluetooth 5 technology arranged sequentially on the perimeter of the analyzed object, each of which contains a transmitting microcontroller and a receiving microcontroller, which are connected via a transmitting microcontroller with a Bluetooth 5 expansion card (4), and this one is connected by serial connection to the central unit (5) connected to the reconstructing server (7). A method for determining the position inside objects using the system. It is characterized in that from the transmitting microcontroller of radio probes (1, 2, 3,..., n) a communication packet with the user data range up to 31 bytes is released, containing the values of RSSI indicators of other scanned devices and the TX power parameter defining the transmit power of the antenna transmitting device, while by means of receiving microcontrollers from radio probes (1, 2, 3,..., n) sent signals are received from other radio probes (1, 2, 3,..., n), from which the receiving microcontrollers calculate RSSI index for the radio probe (1, 2, 3,..., n) from which the packet was sent, and then arranges the RSSI pointers at a fixed location of the data container according to the order number of the radio probe (1, 2, 3,..., n), and all communication packets sent by transmitting microcontrollers are updated by receiving microcontrollers located on the same laminate and are sent out with the help of transmitting microcontrollers from radio probes (1, 2, 3,..., n) working in the Bluetooth technology 5 arranged successively on the perimeter of the analyzed object, a radio signal with a communication packet with a user data range of up to 31 bytes is transmitted, and at irregular intervals it is scanned by the central unit (5), in which after receiving the data, the identification number of the transmitting microcontroller is checked and the received row is placed inside the matrix body in the position consistent with the ordinal number of the radio probe (1, 2, 3,..., n), while irregular intervals between scanning individual system elements cause independent refreshing individual rows of the matrix, while from the central unit (5) a matrix with a dimension of n x n and non-zero diagonal is sent to the reconstructing server (7) using the Apache Kafka data bus, while this data bus does not introduce any changes in the message and without disturbing the data structure and them to algorithms using Tikhonov regularization, Gauss-Newton method, the method of contour sets and artificial intelligence mechanisms, while the reconstruction based on the background matrix and the measurement matrix is presented in the form of a heat map in one of the server services (7).

Description

System and method of positioning inside buildings

The subject of the invention is a system for determining the position inside objects within the framework of radio tomography (RTI) using the Bluetooth 5 standard.

From the patent application CN107911786A there is known a hybrid method of indoor positioning based on the correction of the road network belonging to the field of indoor positioning. The method includes the steps of: mapping and positioning, arranging the base station; acquisition and processing of wireless signals and operation of the hybrid internal positioning algorithm. The method handles various wireless sensor signals including Bluetooth, Wi-Fi, UWB, RFID, and the like. It adopts the method of filtering the wireless signal and determining the area, so irrelevant signals can be further eliminated while the influence of environmental factors on the positioning results can be reduced. A positioning correction method based on interior map data has also been adopted, so that the accuracy of indoor positioning can be further improved. Position computing structure based on position base station and spatial database is adopted, so that position computing performance can be further improved.

Currently, the most widespread indoor location systems are those using the so-called beacons. However, despite the many advantages of networks such as simple implementation, low price or scalability, they are often not able to estimate the user's location with the necessary accuracy. In such systems, it is a superior and critical parameter. The most commonly used protocols for this purpose are technologies using the 2.4 GHz band (ISM - Industrial, Scientific, Medical). Examples include Bluetooth Low Energy technology and technologies related to the Wi-Fi standard.

The subject of the invention is a system and method for determining the position inside objects as part of radio tomography (RTI) using the Bluetooth 5 standard.

The essence of the system is that it consists of radio probes working in Bluetooth 5 technology arranged sequentially on the perimeter of the analyzed object, each of which contains a transmitting microcontroller and a receiving microcontroller, which are connected via a transmitting microcontroller with a Bluetooth 5 expansion card, and this one is connected using a serial connection to the CPU connected to the reconstruction server.

The essence of the method is that from the transmitting microcontroller of radio probes arranged successively on the perimeter of the analyzed object, a communication packet with the user data range up to 31 bytes emerged, containing the values of RSSI indicators of other scanned devices and the TX power parameter defining the transmitting power of the transmitting device antenna. By means of receiving microcontrollers from the radio probes, the signals sent from the remaining radio probes are received, from which the RSSI index for the radio probe from which the packet was sent is calculated by means of receiving microcontrollers. It then arranges the RSSI pointers in a fixed location of the data container according to the order number of the radio probe. All communication packets sent by the transmitting microcontrollers are updated by receiving microcontrollers located on the same laminate and sent out with the help of transmitting microcontrollers from radio probes operating in Bluetooth technology 5 arranged sequentially on the perimeter of the analyzed object a radio signal with a communication packet is transmitted o user data range up to 31 bytes. At irregular intervals, it is scanned by the central unit, in which, after receiving the data, the identification number of the transmitting microcontroller is checked and the received row is placed inside the matrix body in the position corresponding to the order number of the radio probe. Irregular intervals between scanning individual elements of the system cause the individual rows of the matrix to refresh independently. A matrix of n x n dimension, non-zero diagonal, is sent from the CPU to the reconstructing server over the Apache Kafka data bus. This data bus does not introduce any changes to the message and without disturbing the data structure and transmits them to algorithms using Tikhonov regularization, Gauss-Newton method, the method of contour sets and artificial intelligence mechanisms. The reconstruction based on the background matrix and the measurement matrix is presented in the form of a heat map in one of the server services.

The problem to be solved is the use of a commercial communication protocol such as Bluetooth 5 and the development of a method of its application in the radio-tomographic imaging technique. Additional problems include the mutual communication between the radio probe system and the central unit within the same communication protocol. In this case, the system is to be scalable. Of course, it must do it within the permissible load of the radio channels used and the typical range (characteristic) for Bluetooth 5 technology. The approach to the presented issues takes into account the use of computational algorithms and machine intelligence.

A beneficial effect of the use of the system and method for determining the position inside objects is that it can be widely used in any type of building, both in terms of its purpose and method of construction. The proposed solution complies with the protocols and standards for Wi-Fi communication, thus it does not introduce disturbances in wireless communication and can coexist with wireless networks operating on the premises of the facility.

The subject of the invention has been shown in an exemplary embodiment in a schematic drawing.

Fig.1

exemplary embodiment in a schematic drawing.

Examples

The arrangement for determining the position inside the objects presented in the embodiment consists of sixteen radio probes 1, 2, 3, ..., n - Bluetooth 5, each of which includes a transmitting microcontroller and a receiving microcontroller in the form of nRF52832 SoCs. The radio probes 1, 2, 3, ..., n are arranged along the perimeter of the test room according to the ordinal number. The data collected by the receiving part is sent by the transmitting part to the central unit 5 via the Bluetooth 5 expansion card - 4 which makes all the data collected by the operation of the radio probes 1, 2, 3, ..., n available via the serial port. Then, after successfully parsing all the constituent data, the CPU 5 sends it via the Kafka broker 6 to the image reconstruction service located on the server 7.

In the example of the construction of the system and implementation of the method of determining the position inside the objects, the system consists of sixteen radio probes 1, 2, 3, ..., n and a central unit 5, which are located in the test room constituting the office. Due to the irregular shape of the room, the radio probes 1, 2, 3, ..., n have been arranged in such a way as to maintain the highest possible density of the projection angle grid and minimize the insensitivity zones. The central unit 5 has been located halfway from the center of the room. First, the transmitting microcontrollers were calibrated in terms of the transmit power used. For all possible settings, a trial series of background and measurement matrices is made, and then, based on the obtained RSSI indices, the standard deviation is calculated. The smaller the standard deviation, the better the connection quality between the radio probes 1, 2, 3,…, n. In the case of the test room, the power of –4 dBm was the best. Setting too high transmission power may result in the intensification of the reflection of electromagnetic waves, and thus: mutual interference between devices and a decrease in the reliability of measurements. Too little power could cause a sharp increase in the number of unsuccessful transmissions and cause a lack of connection between the most distant points of the system. Another parameter requiring adjustment in the transmitter was the transmission interval, which, if it lasts too long, can slow down the system's work, and when it is too short, it is not necessary to refresh the cells of the measuring matrix with the same values that did not change in the previous cycle. For the selected room, the value has been fixed at 100 ms. In the context of the tracking system, the calibration of the receiving microcontroller also consists in setting two parameters of the Bluetooth scanner: the scanning interval and the width of the scanning window, which is also expressed in a unit of time. The most important criterion when tuning was to set the scanning interval so that it did not occur less frequently than the transmission interval, so that the refresh rate of the results should be at the appropriate level. This parameter is not correlated with the shape and dimensions of the room. The selected microcontrollers have a sensitivity of up to - 100 dBm, so any signal that does not fall below this limit should be successfully received by the scanner module. In the presented system, both the scan interval and the scan window width were set to 50 ms. The standard measurement session involved checking basic scenarios such as: one object in the center of the room, one object on the edge of the room (different points), two objects in the center of the room, two objects on the edge of the room (different points) - in this case the reconstruction was static. For the dynamics study, a well-known object path was established and alternated with calibration to achieve the best results. The most important thing is that the scenario is complementary and moves all the most sensitive points of the measurement zone, such as the center (the highest insensitivity) and points with a reduced density of the projection angle grid. All the described measurements include pairs of matrix groups: background - measurement of at least a dozen pieces in order to average and eliminate sudden RSSI fluctuations.

During the first start-up of the system it is necessary to calibrate it. The process includes creating and entering the room model into the service located on the reconstruction server 7. The model must contain the boundary conditions resulting from the presence of walls, as well as the precise coordinates of each of the installed radio probes 1, 2, 3, ..., n. The model does not require the position of the central unit 5 but it is advisable to be close to the center of the room to standardize the range to each of the radio probes 1, 2, 3, ..., n. This may affect the data refresh rate, in some rows of the matrix.

The methodology of data acquisition and transfer is based on the use of a non-pairing connection type, i.e. advertising, which is made alternately (randomly) on one of the three channels: number 37 at 2402 MHz, number 38 at 2426 MHz and number 39 at 2480 MHz. Each of the radio probes 1, 2, 3, ..., n scans the environment with the receiving microcontroller for transmitting microcontrollers of the remaining radio probes 1, 2, 3, ..., n. Then each packet that is in the buffer and successfully passes the analysis The syntax code is disassembled and the resulting RSSI pointer value is set at a specific point in the line and sent through the serial port to the transmitting part. The position in the line is determined by the ID of the device that represents the pointer. There it is placed in the form of one of the data containers used in broadcasting. The resulting packet is emitted almost immediately, and then captured by successive radio probes 1, 2, 3, ..., n looking for RSSI values. This completes the internal process (only between the radio probes).

Regardless of what processes take place between the radio probes 1, 2, 3, ..., n and their individual parts, the central unit 5 scans the environment in search of the same communication packets but lists completely different data from them, namely complete rows of the matrix to be built. By including two essential radio tomography processes within one Bluetooth 5 package, it becomes possible to limit transactions between devices to 256 analyzes per matrix, which in the case of a sequential system of the same size (sixteen radio probes 1, 2, 3, ..., n) is 900 analyzes. The differences in these values directly affect the speed of the entire location system.

The next step is to load the 256-element matrix into the data publishing function with Apache Kafka. The created data bus is connected directly to the reconstruction server 7 where listening on a given topic is carried out, and all data that is on it are subject to specialized calculations and interpretation.

The location and number of the necessary radio probes 1, 2, 3, ..., n are selected individually for each possible case. Even in the case of two identical rooms (in terms of dimensions), differences in the distribution of EM waves may result from the construction materials used or a different nature (purpose) of the room. The proposed solution complies with the protocols and standards for Wi-Fi communication, thus it does not introduce disturbances in wireless communication and can coexist with wireless networks operating on the premises of the facility. If necessary, the probe system in the scope of radio tomography can be rebuilt many times and easily scalable. It is characterized by easy adaptation to constantly changing working conditions, e.g. the arrangement of furniture in a room. To do this, you only need to wait for the probes to collect the background matrices representing the new layout of the room at rest. If the structure of the room is changed or the placement of the probes is changed, the densities of the created projection angle grid change. In order to counteract the decrease in the reliability of the obtained result, the model of the room located on the server 7 of the reconstructing service should be corrected.

The presented solution can be widely used in any type of building, both in terms of its purpose and construction method. However, it is recommended to pay special attention to public utility buildings, commercial and industrial facilities. This is due to the fact that in this type of facility, the system can additionally support the analysis of human flows and supplement deficiencies in other navigation, security and surveillance systems. It should be emphasized that the system created by the described system does not collect data about a specific user, and thus remains anonymised. It also does not register sensitive data such as image, gender or racial origin, so it complies with the applicable regulations resulting from the implementation of the GDPR provisions.

List of markings:

1 – radio probe number 1
2 – radio probe number 2
3 – radio probe number 3
n – radio probe number n
4 – Bluetooth 5 expansion card
5 – central unit
6 – Kafka broker
7 – Server

Patent Literature

CN107911786A

Claims

A system for determining the position inside objects characterized by the fact that it consists of radio probes (1, 2, 3, ..., n) working in Bluetooth 5 technology arranged sequentially on the perimeter of the analyzed object, each of which contains a transmitting microcontroller and a receiving microcontroller, which are connected via a transmitting microcontroller with a Bluetooth 5 expansion card (4), and this one is connected by serial connection to the central unit (5) connected to the reconstructing server (7).
A method for determining the position inside objects using the system according to claim 1. 1, characterized in that from the transmitting microcontroller of radio probes (1, 2, 3, ..., n) a communication packet with the user data range up to 31 bytes is released, containing the values of RSSI indicators of other scanned devices and the TX power parameter defining the transmit power of the antenna transmitting device, while by means of receiving microcontrollers from radio probes (1, 2, 3, ..., n) sent signals are received from other radio probes (1, 2, 3, ..., n), from which the receiving microcontrollers calculate RSSI index for the radio probe (1, 2, 3, ..., n) from which the packet was sent, and then arranges the RSSI pointers at a fixed location of the data container according to the order number of the radio probe (1, 2, 3, ..., n), and all communication packets sent by transmitting microcontrollers are updated by receiving microcontrollers located on the same laminate and are sent out with the help of transmitting microcontrollers from radio probes (1, 2 , 3, ..., n) working in the Bluetooth technology 5 arranged successively on the perimeter of the analyzed object, a radio signal with a communication packet with a user data range of up to 31 bytes is transmitted, and at irregular intervals it is scanned by the central unit (5), in which after receiving the data, the identification number of the transmitting microcontroller is checked and the received row is placed inside the matrix body in the position consistent with the ordinal number of the radio probe (1, 2, 3, ..., n), while irregular intervals between scanning individual system elements cause independent refreshing individual rows of the matrix, while from the central unit (5) a matrix with a dimension of n x n and non-zero diagonal is sent to the reconstructing server (7) using the Apache Kafka data bus, while this data bus does not introduce any changes in the message and without disturbing the data structure and them to algorithms using Tikhonov regularization, Gauss-Newton method, the method of contour sets and artificial intelligence mechanisms, while the reconstruction based on the background matrix and the measurement matrix is presented in the form of a heat map in one of the server services (7).