Technical Description of Locate-US: an Ultrasonic Local Positioning System based on Encoded Beacons

LOCATE-US is a low-cost ultrasonic local positioning system (U-LPS) for fine-grained indoor positioning of portable devices, such as smartphones or tablets, in order to offer Location Based Services (LBS) such as augmented reality, guidance or multimedia services. It uses encoded signals to obtain a more accurate determination of difference of times of arrival (DTOA), as well as high immunity to noise and robustness against multipath effect and near-far effect. Keywords—LPS, indoor positoning, smartphones, ultrasonic signals. I. GENERAL VIEW OF LOCATE-US SYSTEM A portable ultrasonic beacon architecture has been proposed, so several U-LPS can be easily deployed to cover wide indoor areas. Every U-LPS is formed of five emitters that share roughly the same coverage area, where the receiver to be located is moving around. The emitters in each U-LPS use both, a code division multiple access (CDMA) and a time division multiple access (TDMA) protocols. Every emitter is codified with a different code, with good auto-correlation properties and low mutual interference properties with the others. The emitters for the U-LPS are controlled by a LPC1768 microcontroller [1], that allows the configuration of the ultrasonic transmissions in terms of sampling frequency, modulation schemes and code patterns to be emitted. It also provides a wireless link to an external PC from where the configuration can be easily carried out. At the reception, a non-limited number of receivers can compute its own position by measuring differences of times of arrival (DTOA) of the incoming ultrasonic signals. Thus, there is no need of synchronization between the beacons and the receivers. To overcome audible artifacts and poor resolution (low bandwidth) at the 20-22kHz range available in current smartphones [2], the U-LPS operates around 41kHz and an external hardware device has been included to digitize the incoming signals and send them to the portable device for their processing. It includes a MEMs microphone and a microcontroller that manages the signal digitization and the data sending to the mobile device by means of a USB connection. Then, an Android application at the portable device allows the correlation of the received signals with the code pattern of the emitted codes, computes the DTOAs and the position by means of a hyperbolic trilateration algorithm and represents the trajectory in the device screen in a time less than 0.5s, with a centimetric accuracy. II. BASIC DESCRIPTION OF THE EMITTER ARCHITECTURE Every U-LPS consists of five ultrasonic transmitters, distributed around a 70.7x70.7cm square structure (see Fig. 1). It covers an approximate area of 30m when it is installed on the ceiling at a height of 3.5m. The ultrasonic transmitter is a Prowave 328ST160 [3]. The microcontroller for all the U-LPS provides a wireless link to an external PC, so as to easily configure the sampling frequency, encoding sequences and modulation scheme. These parameters can be modified in runtime, although in practical operation every U-LPS is configured only once in a set-up stage. Specifically, in IPIN’2016 competition, the emitters have been configured to simultaneously transmit Kasami codes [4] with a length of 255 bits. Signals are encoded with the purpose of obtaining a good correlation peak that clearly identifies the time instant of arrival of the signal associated with every beacon, since no synchronization between emission and reception has been used. The encoded signals have been BPSK modulated with two periods of a sinusoidal carrier at 41.67kHz, which has been sampled at 500kHz. Hence, 24 samples (12 in each carrier period) are consider for each bit of the Kasami code. Since only one Digital to Analog Converter (DAC) is available in the microcontroller, a TDMA protocol has been used to transmit the Fig 1.General overview of a portable beacon unit for one U-LPS. Emitters Control unit 2016 International Conference on Indoor Positioning and Indoor Navigation (IPIN), 4-7 October 2016, Alcalá de Henares, Spain codes associated to every emitter at different times. A slot time of 20ms has been devoted for the transmission of every code. Hence, the time interval between two consecutive transmissions of the same code is of 100ms. Considering the competition area in IPIN2016-Track4 of 12mx6m and the ceiling height of 3.56m, three U-LPS are enough to localize the mobile robot that will be moving around if we want good accuracy within all the robot trajectory. The selected Kasami family set provides 16 codes with low crosscorrelation among them, which are sufficient to encode the five beacons of every U-LPS and to ensure low mutual interference and emitter identification even in cases where the robot receive signals from different U-LPSs. Anyway, LOCATE-US can be used in wider areas, as indicated in Fig. 4. To reduce costs, complete U-LPS can be installed only in critical zones, such as entrances or exits, and the inertial sensors of the mobile device together with individual ultrasonic sensors can be used in each zone where the positioning does not need to be so accurate. Bayesian filter algorithms can then be used to combine the information of the U-LPS signals with the inertial sensors of the mobile device. III. BASIC DESCRIPTION OF THE RECEIVER ARCHITECTURE LOCATE-US is thought to offer LBS to users through their mobile phones or tablets. Nevertheless, as in the case of IPINTrack 4 competition, it can also be used to localize mobile robots. A non-limited number of portable receivers can compute their position autonomously by hyperbolic trilateration. To avoid frequency limitations and restrictions due to the sampling rate of current smartphones (44.1kHz), a low cost ultrasonic signal acquisition module has been designed (see [5] for further details). Fig. 5 shows the block diagram of the proposed adaptation hardware, while Fig. 6 shows a picture of a real device. Note its reduced dimensions and ease of connection to the portable device. The acquisition module includes a MEMS SPU0414HR5H-SB [6] microphone with an adequate bandwidth around the 41.67kHz. The incoming signals captured by the microphone are then high-pass filtered and amplified (or attenuated) through a programmable gain module controlled by a STM32F103. Thus, the gain can be adjusted depending on the application requirements and signal level. The microcontroller also digitizes the incoming signals at a sampling rate of 100kHz. A decimation by 5 is carried out to reduce the amount of data to be processed. Due to memory restrictions in the microcontroller, a buffer with 13.000 samples has been considered. Once the buffer is filled, it is send over a USB link to the portable device and a new acquisition starts. Hence 0.13s of the ultrasonic signal has to be saved to be processed. It assures that at least one complete period of every beacon code is received. Once in the device, an Android application is run to obtain the location and depict the trajectory in the mobile screen. The application includes all the low level processing (signal demodulation and correlation, peak detection and TDOA computing algorithms), as well as the high level processing (position estimation through hyperbolic trilateration, map Fig 2.Block diagram of the emitter control unit. Fig 3.Photo of the emitter control unit. 255 bits Kasami BPSK Modulated Codes DAC LPC1768 (Cortex M-3 microcontroller)