Predictive analysis on mobile network traffic is becoming of fundamental importance for the next generation cellular network. Proactively knowing the user demands, allows the system for an optimal resource allocation. In this paper, we study the mobile traffic of an LTE base station and we design a system for the traffic prediction using Recurrent Neural Networks. The mobile traffic information is gathered from the Physical Downlink Control CHannel (PDCCH) of the LTE using the passive tool presented in [1]. Using this tool we are able to collect all the control information at 1 ms resolution from the base station. This information comprises the resource blocks, the transport block size and the modulation scheme assigned to each user connected to the eNodeB. The design of the prediction system includes long short term memory units. With respect to a Multilayer Perceptron Network, or other artificial neurons structures, recurrent networks are advantageous for problems with sequential data (e.g. language modeling) [2]. In our case, we state the problem as a supervised multivariate prediction of the mobile traffic, where the objective is to minimize the prediction error given the information extracted from the PDCCH. We evaluate the one-step prediction and the long-term prediction errors of the proposed methodology, considering different numbers for the duration of the observed values, which determines the memory length of the LSTM network and how much information must be stored for a precise traffic prediction.

Cashiers in retail stores usually exhibit certain repetitive and periodic activities when processing items. Detecting such activities plays a key role in most retail fraud detection systems. In this paper, we propose a highly efficient, effective and robust vision technique to detect checkout-related primitive activities, based on a hierarchical finite state machine (FSM). Our deterministic approach uses visual features and prior spatial constraints on the hand motion to capture particular motion patterns performed in primitive activities. We also apply our approach to the problem of retail fraud detection. Experimental results on a large set of video data captured from retail stores show that our approach, while much simpler and faster, achieves significantly better results than state-of-the-art machine learning-based techniques both in detecting checkout-related activities and in detecting checkoutrelated fraudulent incidents.

Among the smart capabilities promised by the next generation cellular networks (5G and beyond), it is fundamental that potential network anomalies are detected and timely treated to avoid critical issues concerning network performance, security, public safety. In this paper, we propose a comprehensive framework for detecting network anomalies using mobile traffic data: collecting data from the LTE Physical Downlink Control Channel (PDCCH) of different eNodeBs, we implement deep learning algorithms in a semi-supervised way to detect potential traffic anomalies that are generated, for example, by unexpected crowd gathering. With respect to other types of mobile dataset, using LTE PDCCH information, we are able to obtain fine-grained and high-resolution data for the users that are connected to the LTE eNodeB. Through a semi-supervised approach, algorithms are trained to detect anomalies using only one class of traffic samples. We design two algorithms based on stacked-LSTM Neural Networks: 1) LSTM Autoencoder (LSTM-AE), in which the objective is to reconstruct the traffic samples 2) LSTM traffic predictor (LSTM-PRED), where the goal is to predict the traffic in the next time-instants, based on historical data. In both cases, we analyze the reconstruction (or prediction) error to assess if the mobile traffic presents anomalies or not. Using the F1-score as metric, we demonstrate that the proposed methods are able to identify the anomalous traffic periods, beating a benchmark that comprises different state-of-the-art algorithms for anomaly detection.

For safety purposes, railroad tracks need to be inspected on a regular basis for physical defects or design noncompliances. Such track defects and non-compliances, if not detected in a timely manner, may eventually lead to grave consequences such as train derailments.In this paper, we present a real-time automatic visionbased rail inspection system, with main focus on anchorsan important rail component type, and anchor-related rail defects, or exceptions. Our system robustly detects important rail components including ties, tie plates, anchors with high accuracy and efficiency. Detected objects are then consolidated across video frames and across camera views to map to physical rail objects, by combining the video data streams from all camera views with GPS information and speed information from the distance measuring instrument (DMI). After these rail components are detected and consolidated, further data integration and analysis is followed to detect sequence-level track defects, or exceptions. Quantitative analysis performed on a real online field test conducted on different track conditions demonstrates that our system achieves very promising performance in terms of rail component detection, anchor condition assessment, and compliance-level exception detection. We also show that our system outperforms another advanced rail inspection system in anchor detection.

In this paper, we present a real-time automatic vision-based rail inspection system, which performs inspections at 16 km/h with a frame rate of 20 fps. The system robustly detects important rail components such as ties, tie plates, and anchors, with high accuracy and efficiency. To achieve this goal, we first develop a set of image and video analytics and then propose a novel global optimization framework to combine evidence from multiple cameras, Global Positioning System, and distance measurement instrument to further improve the detection performance. Moreover, as the anchor is an important type of rail fastener, we have thus advanced the effort to detect anchor exceptions, which includes assessing the anchor conditions at the tie level and identifying anchor pattern exceptions at the compliance level. Quantitative analysis performed on a large video data set captured with different track and lighting conditions, as well as on a realtime field test, has demonstrated very encouraging performance on both rail component detection and anchor exception detection.Specifically, an average of 94.67% precision and 93% recall rate has been achieved for detecting all three rail components, and a 100% detection rate is achieved for compliance-level anchor exception with three false positives per hour. To our best knowledge, our system is the first to address and solve both component and exception detection problems in this rail inspection area.