User Orientation Detection in Relation to Antenna Geometry in Ultra-Wideband Wireless Body Area Networks Using Deep Learning.

In this paper, the issue of detecting a user's position in relation to the antenna geometry in ultra-wideband (UWB) off-body wireless body area network (WBAN) communication using deep learning methods is presented. To measure the impulse response of the channel, a measurement stand consisting of EVB1000 devices and DW1000 radio modules was developed and indoor static measurement scenarios were performed. It was proven that for the binary classification of user orientation, neural networks achieved accuracy that was more than 9% higher than that for the well-known threshold method. In addition, the classification of user position angles relative to the reference node was analyzed. It was proven that, using the proposed deep learning approach and the channel impulse response, it was possible to estimate the angle of the user's position in relation to the antenna geometry. Absolute user orientation angle errors of about 4-7° for convolutional neural networks and of about 14-15° for multilayer perceptrons were achieved in approximately 85% of the cases in both tested scenarios.

Channel State Estimation in LTE-Based Heterogenous Networks Using Deep Learning.

Following the continuous development of the information technology, the concept of dense urban networks has evolved as well. The powerful tools, like machine learning, break new ground in smart network and interface design. In this paper the concept of using deep learning for estimating the radio channel parameters of the LTE (Long Term Evolution) radio interface is presented. It was proved that the deep learning approach provides a significant gain (almost 40%) with 10.7% compared to the linear model with the lowest RMSE (Root Mean Squared Error) 17.01%. The solution can be adopted as a part of the data allocation algorithm implemented in the telemetry devices equipped with the 4G radio interface, or, after the adjustment, the NB-IoT (Narrowband Internet of Things), to maximize the reliability of the services in harsh indoor or urban environments. Presented results also prove the existence of the inverse proportional dependence between the number of hidden layers and the number of historical samples in terms of the obtained RMSE. The increase of the historical data memory allows using models with fewer hidden layers while maintaining a comparable RMSE value for each scenario, which reduces the total computational cost.

Software-Defined NB-IoT Uplink Framework-The Design, Implementation and Use Cases.

In the radiocommunication area, we may observe a rapid growth of new technology, such as 5G. Moreover, all the newly introduced radio interfaces, e.g., narrowband Internet of Things (NB-IoT), are strongly dependent on the software. Hence, the radiocommunication software development and optimization, as well as the 3GPP technical specification, should be introduced at the academic level of education. In this paper, a software-defined NB-IoT uplink framework in the field of design is presented, as well as its realization and potential use cases. The framework may be used as an academic tool for developing, investigating, and optimizing the digital transmitter paths. The proposed realization is focused on the key elements in the physical layer of the NB-IoT interface used in the sensor devices. Furthermore, the paper also highlights the need of the data processing optimization to minimize the power consumption and usage of the resources of the NB-IoT node during transmitting gathered telemetric data.

Person Tracking in Ultra-Wide Band Hybrid Localization System Using Reduced Number of Reference Nodes.

In this article a novel method of positional data integration in an indoor hybrid localization system combining inertial navigation with radio distance measurements is presented. A point of interest is the situation when the positional data and the radio distance measurements are obtained from less than thee reference nodes and it is impossible to unambiguously localize the moving person due to undetermined set of positional equations. The presented method allows to continuously provide localization service even in areas with disturbed propagation of the radio signals. Authors performed simulation and measurement studies of the proposed method to verify the precision of position estimation of a moving person in an indoor environment. It is worth noting that to determine the simulation parameters and realize the experimental studies the hybrid localization system demonstrator was developed, combining inertial navigation and radio distance measurements. In the proposed solution, results of distance measurements taken to less than three reference nodes are used to compensate the drift of the position estimated using the inertial sensor. In the obtained simulation and experimental results it was possible to reduce the localization error by nearly 50% regarding the case when only inertial navigation was used, additionally keeping the long term root mean square error at the level of ca. 0.50 m. That gives a degradation of localization precision below 0.1 m with respect to the fusion Kalman filtration when four reference nodes are present.

Deep Learning-Based LOS and NLOS Identification in Wireless Body Area Networks.

In this article, the usage of deep learning (DL) in ultra-wideband (UWB) Wireless Body Area Networks (WBANs) is presented. The developed approach, using channel impulse response, allows higher efficiency in identifying the direct visibility conditions between nodes in off-body communication with comparison to the methods described in the literature. The effectiveness of the proposed deep feedforward neural network was checked on the basis of the measurement data for dynamic scenarios in an indoor environment. The obtained results clearly prove the validity of the proposed DL approach in the UWB WBANs and high (over 98.6% for most cases) efficiency for LOS and NLOS conditions classification.