Analyzing the code structures of multidimensional signals for a continuous information transmission channel

One of the directions to improve the efficiency of modern telecommunication systems is the transition to the use of multidimensional signals for continuous channels of information transmission. As a result of studies carried out in recent years, it has been established that it is possible to ensure high quality of information transmission in continuous channels by combining demodulation and decoding operations into a single procedure that involves the construction of a code construct for a multidimensional signal. This paper considers issues related to estimating the possibility to improve the efficiency of continuous information transmission channel by changing the signal distance of the code structure. It has been established that the code structures of such types as a hierarchical code construct of signals, a hierarchical code construct of signals with Euclidean metric, a reversible code construct of signals, a reversible code construct of signals with Euclidean metric have the potential, when used, to increase the speed of information transmission along a continuous channel. With a signal distance reduced by 10 percent or larger, it could increase by two times or faster. The estimation of the effect of reducing a signal distance on the efficiency of certain types of code structures was carried out. It has been established that the hierarchical reversible code construct, compared to the hierarchical code construct, provides a win of up to two or more times in the speed of information transmission with a halved signal distance. Implementing the modulation procedure has no fundamental difficulties, on the condition that for each code of the code construct the encoding procedure is known when using binary codes. The results reported here make it possible to build an acceptably complex demodulation procedure according to the specified types of code structures

Shedding Light on People Action Recognition in Social Robotics by Means of Common Spatial Patterns

Action recognition in robotics is a research field that has gained momentum in recent years. In this work, a video activity recognition method is presented, which has the ultimate goal of endowing a robot with action recognition capabilities for a more natural social interaction. The application of Common Spatial Patterns (CSP), a signal processing approach widely used in electroencephalography (EEG), is presented in a novel manner to be used in activity recognition in videos taken by a humanoid robot. A sequence of skeleton data is considered as a multidimensional signal and filtered according to the CSP algorithm. Then, characteristics extracted from these filtered data are used as features for a classifier. A database with 46 individuals performing six different actions has been created to test the proposed method. The CSP-based method along with a Linear Discriminant Analysis (LDA) classifier has been compared to a Long Short-Term Memory (LSTM) neural network, showing that the former obtains similar or better results than the latter, while being simpler.