All industries are trying to leverage Artificial Intelligence (AI) based on their existing big data which is available in so called tabular form, where each record is composed of a number of heterogeneous continuous and categorical columns also known as features. Deep Learning (DL) has constituted a major breakthrough for AI in fields related to human skills like natural language processing, but its applicability to tabular data has been more challenging. More classical Machine Learning (ML) models like tree-based ensemble ones usually perform better. This paper presents a novel DL model using Graph Neural Network (GNN) more specifically Interaction Network (IN), for contextual embedding and modeling interactions among tabular features. Its results outperform those of a recently published survey with DL benchmark based on seven public datasets, also achieving competitive results when compared to boosted-tree solutions.

IoT networks connect dispersed and heterogeneous devices with high economic impact, making them an important cybersecurity target. The changing nature of cybersecurity attacks requires intrusion detectors with the ability to detect new attacks. Our proposed model is a network intrusion detector created specifically for the needs of IoT networks with the aim of detecting previously unseen attacks. It is based on a shallow neural network (NN) architecture following a novel contrastive learning scheme. In this scheme, both the network features and the labels are projected into a common representation (embedding) space where a similarity score is defined. The labels act as a prototype for each type of traffic in embedding space, and classification is based on the proximity of samples to these class prototypes. The dimensionality and structure of the embedding space are critical. In this work, we explore the advantages of having an embedding space with expanded dimensionality using a kernel approximation technique (Random Fourier Features) that is integrated and learned within the NN. To avoid overfitting, we investigate the importance of various regularization techniques (L2 and contractive). The resulting model is tested against three network intrusion detection data sets to assess its ability to detect known and unknown attacks (zero-shot learning). The experimental results show a higher ability of the proposed model to detect unknown attacks than similar models and alternative machine learning models, in the literature.

Internet of Things (IoT) can be combined with machine learning in order to provide intelligent applications to the network nodes. Furthermore, IoT expands these advantages and technologies to the industry. In this paper, we propose a modification of one of the most popular algorithms for feature selection, fast-based-correlation feature (FCBF). The key idea is to split the feature space in fragments with the same size. By introducing this division, we can improve the correlation and, therefore, the machine learning applications that are operating on each node. This kind of IoT applications for industry allows us to separate and prioritize the sensor data from the multimedia-related traffic. With this separation, the sensors are able to detect efficiently emergency situations and avoid both material and human damage. The results show the performance of the three FCBF-based algorithms for different problems and different classifiers, confirming the improvements achieved by our approach in terms of model accuracy and execution time.

Legacy networks, both fixed and mobile, which were originally designed for voice communications, are progressively migrating to new infrastructures that promise to revolutionize the services offered. In this paper, we will cover this new generation of personal communication services, with an emphasis on the family of Internet protocol (IP)-based multimedia subsystem (IMS)-aided infrastructure that relies on the session initiation protocol (SIP). As a benefit, the end users will enjoy a new generation of personal communications services that are accessible anywhere and anytime. These services are directly related to the end users rather than to their diverse devices. It is anticipated that the new deployments of next-generation networks (all-IP based) will accelerate the adoption of the IMS technology.

UMTS Services Accessibility is a daily concern for network operators. Proper network resource dimensioning allows maximizing the operator investment and enhances user experience. In the case of a resource shortage, it is very helpful to decide which resource is mainly responsible for the downgrade, in order to upgrade this particular resource. In this article, authors model UTRAN accessibility, once known the system resources, the different service''s resource consumptions and each service''s traffic. The model is validated through comparison between theoretical and real accessibility, and an accuracy enhancement is achieved using loopback correction in several real case studies. Results show that the proposed model matches the binary decision of degraded or non- degraded node with errors below 10% and it is able to suggest the most congested resource each time.

This paper describes the architecture of a proposed multiparty conferencing system for satellites. Different conferencing models are discussed and compared. A session initiation protocol (SIP)-based conference signaling model and an extension to protocol-independent multicast sparse mode that supports quality-of-service (QoS) in DiffServ networks are proposed, as particularly suitable for multiparty conferencing applications over satellite links. The paper also presents key issues and potential solutions of scalable QoS multicast services for multiparty conferences over satellite. End-to-end QoS parameters for voice and video are measured and analyzed on a prototype.

Future broadband satellite access networks will provide user connectivity via a return channel from a centralized hub. The forward channel can be embedded into the broadcast channel yielding asymmetric bandwidth ratios of 10:1 or greater. To meet such a bandwidth asymmetry as well as the required quality of service, QoS, selection of multiple access and media access control protocols plays a significant role among other design issues. In most of the recent satellite access networks, the forward traffic is broadcasted to satellite interactive terminals (SITs) using the ETSI specified DVB-S protocol. For the return channel multiple access protocol, options based on multi-frequency time division multiple access (MF-TDMA), code division multiple access (CDMA) and random multiple access (RMA) are available. We present a comparative overview of connection- and contention-oriented multiple access and focus on the use of spread ALOHA by using one single long code.

There are many problems regarding the performance of the return channel of a hybrid fiber-coaxial (HFC) network, as cable operators know well. These problems are mainly due to the branch-tree architecture of the coaxial part of the network, which causes all the noise from any point to add in the upstream direction, an effect called noise funneling. The present work has focused on the controlled insertion of selected types of noise into a part of a coaxial network, simultaneously with the data transmission of a cable modem towards the head end. As a result, the upstream (return) data transmission will be affected, depending on the various changing and controlled noise characteristics. Monitoring several parameters shows the effect on the transmitted signal, thus giving an idea of the quality of the transmission.