Graph Neural Networks (GNNs) have exploded onto the machine learning scene in recent years owing to their capability to model and learn from graph-structured data. Such an ability has strong implications in a wide variety of fields whose data are inherently relational, for which conventional neural networks do not perform well. Indeed, as recent reviews can attest, research in the area of GNNs has grown rapidly and has lead to the development of a variety of GNN algorithm variants as well as to the exploration of ground-breaking applications in chemistry, neurology, electronics, or communication networks, among others. At the current stage research, however, the efficient processing of GNNs is still an open challenge for several reasons. Besides of their novelty, GNNs are hard to compute due to their dependence on the input graph, their combination of dense and very sparse operations, or the need to scale to huge graphs in some applications. In this context, this article aims to make two main contributions. On the one hand, a review of the field of GNNs is presented from the perspective of computing. This includes a brief tutorial on the GNN fundamentals, an overview of the evolution of the field in the last decade, and a summary of operations carried out in the multiple phases of different GNN algorithm variants. On the other hand, an in-depth analysis of current software and hardware acceleration schemes is provided, from which a hardware-software, graph-aware, and communication-centric vision for GNN accelerators is distilled.

No abstract

The reconfigurable intelligent surface is a promising technology for the manipulation and control of wireless electromagnetic signals. In particular, it has the potential to provide significant performance improvements for wireless networks. However, to do so, a proper reconfiguration of the reflection coefficients of unit cells is required, which often leads to complex and expensive devices. To amortize the cost, one may share the system resources among multiple transmitters and receivers. In this paper, we propose an efficient reconfiguration technique providing control over multiple beams independently. Compared to time-consuming optimization techniques, the proposed strategy utilizes an analytical method to configure the surface for multibeam radiation. This method is easy to implement, effective and efficient since it only requires phase reconfiguration. We analyze the performance for indoor and outdoor scenarios, given the broadcast mode of operation. The aforesaid scenarios encompass some of the most challenging scenarios that wireless networks encounter. We show that our proposed technique provisions sufficient improvements in the observed channel capacity when the receivers are close to the surface in the indoor office environment scenario. Further, we report a considerable increase in the system throughput given the outdoor environment.

No abstract

Abstract-The future wireless networks are expected to be extremely dense and heterogeneous, with the users experiencing multi-connectivity through the multiple available radio access technologies (RATs). These prevalent characteristics, along with the strict QoS requirements, renders the handover (HO) process optimization as a critical objective for future networks. Along side the evolving network characteristics and methodologies, an evolving network architecture needs to be considered as well. Such evolution should not only facilitate HO process enhancement, i.e., reduction in HO delay and signaling, but it should also allow for a smooth transition from current to future wireless networks. Hence, in this work we firstly present an evolutionary core network entity called the Integrated MME-SDN Controller and the associated network architecture. The proposed architecture provides a migratory path for the existing 3GPP cellular architectures towards the 5G networks. Next, we discuss the benefits and challenges of such an architectural approach, with one of the benefits being a manageable CAPEX for the network operators through its transitional nature. Subsequently, utilizing the aforementioned proposed architecture, we present the handover process enhancement for the current 3GPP defined HO processes. We quantify the improvements achieved in terms of latency, transmission and processing cost for the different 3GPP HO processes. We also show that the proposed HO mechanism leads to a significant reduction in latency and signaling for certain types of HOs which, as a consequence, will critically benefit any dense and heterogeneous wireless system, such as 5G.

Mobility management is a critical component for any new wireless standard to be ubiquitous. While 4G-LTE and prior wireless standards utilized vendor specific hardware and software on which mobility management (MM) functionality was implemented, recent 5G architecture releases by 3GPP indicate a complete departure from the same. 3GPP in release 14 and the upcoming release 15 has stressed upon the utilization of Software Defined Networking (SDN) and Network Function Virtualization (NFV) as the drivers of 5G technology. Consequently, new challenges related to MM and specifically handover management will be encountered owing to the inter-working setup between the 5G Next-Gen Core (NGC) and the Evolved Packet System (EPS) core. In this paper, we exploit the SDN to enhance the signaling of the HO methods proposed by 3GPP. Although the proposed approach can be applied to any HO method, in this paper we specifically evaluate the scenario wherein a dedicated interface between the Mobility Management Entity (MME) in the EPC and the Access and Mobility management Function (AMF) in the 5G NGC, i.e., N26 as specified by 3GPP, is nonexistent. Such a scenario is reasonable during the initial deployment phases of 5G networks. We show that the proposed mechanism is efficient as compared to the 3GPP handover strategy in terms of latency, transmission and processing costs.

In this article, we articulate the technical challenges to enable a future AR/VR end-to-end architecture, that combines 5G URLLC and Tactile IoT technology to support this next generation of interconnected AR/VR applications. Through the use of IoT sensors and actuators, AR/VR applications will be aware of the environmental and user context, supporting human-centric adaptations of the application logic, and lifelike interactions with the virtual environment. We present potential use cases and the required technological building blocks. For each of them, we delve into the current state of the art and challenges that need to be addressed before the dream of remote AR/VR interaction can become reality.

No abstract