The fingerprinting technique is a popular approach to reveal location of persons, instruments or devices in an indoor environment. Typically based on signal strength measurement, a power level map is created first in the learning phase to align with measured values in the inference. Second, the location is determined by taking the point for which the recorded received power level is closest to the power level actually measured. The biggest limit of this technique is the reliability of power measurements, which may lack accuracy in many wireless systems. To this end, this work extends the power level measurement by using multiple anchors and multiple radio channels and, consequently, considers different approaches to aligning the actual measurements with the recorded values. The dataset is available online. This article focuses on the very popular radio technology Bluetooth Low Energy to explore the possible improvement of the system accuracy through different machine learning approaches. It shows how the accuracy–complexity trade-off influences the possible candidate algorithms on an example of three-channel Bluetooth received signal strength based fingerprinting in a one dimensional environment with four static anchors and in a two dimensional environment with the same set of anchors. We provide a literature survey to identify the machine learning algorithms applied in the literature to show that the studies available can not be compared directly. Then, we implement and analyze the performance of four most popular supervised learning techniques, namely k Nearest Neighbors, Support Vector Machines, Random Forest, and Artificial Neural Network. In our scenario, the most promising machine learning technique being the Random Forest with classification accuracy over 99%.

In the evolving landscape of Industry 4.0, the integration of advanced wireless technologies into manufacturing processes holds the promise of unprecedented connectivity and efficiency. In particular, the data transmission in a heavy industry environment needs stable connectivity with mobile operators. This paper deals with the performance study of 4G and 5G mobile signal coverage within a complex factory environment. For this purpose, a cost-effective and portable measurement setup was realized and used to provide long-term measurement campaigns monitoring and recording several key parameter indicators (KPIs) in 4G/5G downlink and upload. To support the reproducibility of the provided study and other research activities, the measured dataset is publicly available for download. Among others findings, the obtained results show how the performance of 4G/5G is influenced by a heavy industry environment and of the time of day on the network load.

Indoor positioning and navigation increasingly has become popular and there are many different approaches, using different technologies. In nearly all of the approaches the locational accuracy depends on signal propagation characteristics of the environment. What makes many of these approaches similar is the requirement of creating a signal propagation Radio Map (RM) by analysing the environment. As this is usually done on a regular grid, the collection of Received Signal Strength Indicator (RSSI) data at every Reference Point (RP) of a RM is a time consuming task. With indoor positioning being in the focus of the research community, the reduction in time required for collection of RMs is very useful as it allows researchers to spend more time with research instead of data collection. In this paper we analyse the options for reducing the time required for the acquisition of RSSI information. We approach this by collecting initial RMs of Wi-Fi signal strength using 5 ESP32 micro controllers working in monitoring mode and placed around our office. We then analyse the influence the approximation of RSSI values in unreachable places has, by using linear interpolation and Gaussian Process Regression (GPR) to find balance between final positioning accuracy, computing complexity, and time requirements for the initial data collection. We conclude that the computational requirements can be significantly lowered, while not affecting the positioning error, by using RM with a single sample per RP generated considering many measurements.

The paper presents a study of existing methods for motion and face detection algorithms and their application to the on-board miniature Raspberry Pi computer. The algorithms realized by OpenCV functions were modified to optimize their operation on the mentioned platform, which could be used as an embedded surveillance system. The paper also mentions the training of a custom classifier for hand detection, what could be further used as a basis for detecting hand gestures.

Paper presents the optimized implementation of the digital signal processing algorithms (real and complex Fast Fourier Transforms) for the specific hardware architecture. The algorithms' source codes were optimized at low level, while all redundant operations (e.g. branching instructions) were avoided. Contrary to results compiled from the high level codes, time consuming load/store operations were considerably eliminated as well and temporal data were stored in the general purpose registers. Contrary to other implementations, the several calls of the identical functions (but with shared data) provide a reducing of the processor idle states. The TMS320C6748 and TMS320C6678 digital signal processors with the Very Long Instruction Word architecture were used for the implementation of proposed functions. The average duration of FFT optimized functions is between five CPU cycles for four real values and 44 CPU cycles for sixteen real values, respectively.

This paper describes tests performed by parallel signal processing algorithms on Texas Instruments 8-core digital signal processor and Intel Core i7. Tests reveal how the change of number of threads and size of input data influences the performance compared with sequential code. The result of the test is that the relative speedup is highly dependent on type of the algorithm and the amount of processed data.