MOIN (Minimum Location Infrastructure)
|Duration:||01.09.2018 - 31.12.2020|
||Prof. Dr.-Ing Horst Hellbrück|
|Staff:||Sven Ole Schmidt, Wiland Arlt|
A producer stores his semi-finished and pre-products in a hall with numerous narrow aisles in between. Inside the production hall, there are also completely enclosed rooms (sanding, welding and painting booths). From semi-finished and pre-products, the final product is manufactured in several operations. In addition to retrieving the semi-finished and precursors, the entire production process should be automatically monitored, e.g. Determine buffer times. In order to perform an automatic tracking of the production process, the position of the semi-finished and pre-products, as well as the production order, must be retrievable at any time.
With traditional methods, such as barcode or passive RFID, complete tracking is only possible to a limited extent. In addition, these methods cause considerable manual overhead (permanent simultaneous scanning of object and position) and pose a high potential for errors by forgetting a scan. The object can no longer be located in this case.
Existing positioning systems based on radio runtime measurement require a large number of reference points. Due to the hall topology described, the corresponding infrastructure for radiolocation would have to be set up in each individual aisle in this case. A location within the enclosed spaces would be limited or not possible by shielding.
The aim of this project is the redevelopment of a localization system, which reduces the contained infrastructure to a minimum. For this purpose, the multipath propagation, which is caused by the wireless transmission, is used constructively to initialize virtual anchors from a given anchor, while taking into account the hall topology. Then, these are used analogous to real anchors, which reduces the hardware requirements.
Figure: Minimization of hardware in an indoor localization system
In the subproject Modeling and Algorithms (ModAl) of this cooperation project, the University of Applied Sciences Lübeck is developing a simulation and planning tool that can analyze a given spatial geometry and perfectly position the localization anchor on the basis of algorithmics. In consideration of this planning, we also develop an algorithmic position analysis of the tags.
An important tool for tag localization is the channel impulse response. This describes the multipath propagation through a diagram which maps the received signal in the form of the received signal copies with their individual power and transmission duration.
Figure: Measurement of a channel impulse response
By correlating with the different paths of multipath propagation, a clear inference of the tag position can be drawn from the received channel impulse response.
|||MAMPI – Multipath-assisted Device-free Localization with Magnitude and Phase Information , In International Conference on Localization and GNSS, 2020. [bib] [abstract]|
Narrowband radio-frequency (RF)-based device-free localization (DFL) systems suffer from multipath propagation. Therefore, we propose multipath-assisted device-free localization with magnitude and phase information (MAMPI), a DFL system that employs ultra-wideband (UWB) channel impulse response (CIR) measurements between a transmitter and receiver and thereby benefits from multipath propagation. We implement a model that includes raytracing, diffraction, and inaccuracies of measurements in a validated error model. The multipath-assisted DFL system creates feature vectors at any position based on magnitude and phase measurements of each multipath component (MPC) of the CIR. We evaluate our new approach with a position error probability based on the propagation model including errors, as well as a distance metric of the feature vector of the positions. We compare the performance of the system with a line-of-sight solution with four instead of two nodes and variants of magnitude-only and phase-only approaches. By combining the magnitude and phase measurements of a multipath-assisted DFL system, we achieve a position error probability that is similar to the conventional DFL system.
|||On the Effective Length of Channel Impulse Responses in UWB Single Anchor Localization , In International Conference on Localization and GNSS, 2019. [bib] [abstract]|
Recently, single anchor localization evolves as a new research topic that exploits multipath propagation for calculation of tag positions. With a combination of movement information and particle filters, they provide a precision that is similar to multi-anchor systems. However, a systematic approach to the design and implementation of such systems is not yet available. The combination of theory and mathematical modeling for channel impulse responses is still an open research question that we address in this paper. Therefore, we propose a new representation of a channel impulse response targeted for single anchor localization systems. Based on this representation, we model the relationship between tag positions and channel impulse responses and evaluate the statistic properties of channel impulse responses in this application. In this paper, we introduce a new metric for the assessment of anchor positions, the effective length of CIRs. By the shortest effective length of a set of CIRs, we identify the best anchor position, since it indicates the position where requirements for the measurement of the channel impulse response are lowest.
|||Modeling the Magnitude and Phase of Multipath UWB Signals for the Use in Passive Localization , In 16th Workshop on Positioning, Navigation and Communication, 2019. [bib] [abstract]|
Radio-frequency (RF)-based device-free localization (DFL) systems measure RF parameters such as the received signal strength or channel state information to detect and track objects within a certain area. However, the change of the RF signal caused by the object is superimposed with various changes of the RF signal due to multipath propagation, especially in indoor environments. In this paper, we develop a model for ultra-wideband (UWB) channel impulse response (CIR) measurements for application in DFL systems. The model predicts received signal parameters in a setup with a transmitter and a receiver node, a person and multipath propagation. Different from other approaches, the RF hardware, and the model provides both magnitude and phase information for individual multipath components. We evaluate the new model with real measurements that have been conducted with a Decawave DW1000 radio chip. For the magnitudes, we achieved a correlation factor from 0.78 to 0.87 and maximum mean and standard deviation errors of 1.7 dB and 2.2 dB respectively. For the phase, we achieved correlation factor from 0.6 to 0.81 and maximum mean and standard deviation errors of 0.32 dB and 0.47 dB respectively, showing that the prediction of our proposed model for the magnitude and phase fits well to our measurements.
|||Understanding and Prediction of Ultra-Wide Band Channel Impulse Response Measurements , Technical report, Technische Universität Braunschweig, 2019. [bib] [pdf] [abstract]|
Recently, ultra-wide band transceiver systems have provided data transfer, timestamps and channel impulse response measurements to the user. The interpretation of the timestamps and the channel impulse response, however, is difficult and not intuitive. In simple scenarios, line of sight and non-line of sight pulses can be distinguished easily, which simplifies the reconstruction. For more complex scenarios, the interpretation remains difficult and is still an unsolved problem. In this paper, we investigate the channel impulse response measurements of the DecaWave DW1000 ultra-wide band transceiver and model the expected results for simple scenarios based on information provided from the transceiver data sheet. We will show that we are able to predict the measurement results of the transceiver with acceptable accuracy by applying the model above in experiments.
|||Improvements to UWB Channel Impulse Response Measurements for Indoor Localization , (T. M. Buzug et. al., ed.), 2019. [bib] [abstract]|
This paper proposes two methods to improve the accuracy of UWB channel impulse response (CIR) measurements. Improving the accuracy of UWB CIR measurements results in an improvement of the overall accuracy of an indoor localization system. Two methods are analyzed, both with the idea of combining a series of CIR measurements to yield higher accuracy than a single measurement. We evaluated both methods by gathering data at different positions within a room. These methods reduced the error by an average of 7% and 4% respectively. The results indicate that utilizing these techniques will improve the accuracy of localization.