Industrial IoT(IIoT) and M2M Networks.
In this research line we deeply study Industrial networks and contribute to their standardization. Industrial Networks have developed alongside the traditional Internet. This is because an industrial network (an “Operational Technology”– OT) has requirements very different from the Internet (an “Information Technology” – IT). While the Internet is built to interconnect billions of heterogeneous devices communicating large amounts of data over a long distance, an industrial network is typically deployed within a factory floor, typically connecting 100’s or 1,000’s of devices. And although the amount of data in typical industrial process applications may not be large, what is critical is reliability (all data is received by its final destination), latency (using guaranteed time bounds, as opposed to best-effort) and battery lifetime. Popular wireless link layers (based on Carrier Sense Multiple Access – CSMA) do not meet these expectations; thus, industrial networks have remained traditionally wired. The cost and the operative limitations of wired networks is triggering a new generation of wireless communication technologies and ongoing standards which shift the connectivity paradigm in industries as operative costs are drastically reduced. WINE focuses from a practical perspective in the convergence of these technologies to the current Internet infrastructure (IP enabled), including routing, scheduling, energy efficiency, resource sharing and optimization techniques to develop the basis for communications of the Industrial Internet of Things.
Contextual intelligence in the Internet of Everything (IoE)
Ubiquitous computing together with wireless networking is enabling the Internet of Everything, where information systems, people, and a wide variety of objects are becoming seamlessly interconnected. However, bringing objects to the Internet presents challenges regarding data throughput, number of devices, power consumption or read range. Most of these challenges are linked with the context where the communication devices are placed, and thus, “understanding” that context is key for the IoE performance improvement. Yet, the information obtained in IoE scenarios can be exploited in combination of combinatorial optimization techniques to facilitate the development and understanding of complex dynamics in real environments such as cities, roads or dynamic systems in general
Around this concept, this research line aims to extract context-aware information to improve “Internet of Everything” networking technologies such as RFID, 802.15.4 or other low power networks. By using information provided by sensors or radio frequency parameters, machine or other related learning techniques will be used to address the current IoE challenges. The goal is to exploit contextual intelligence (in an individual or collaborative basis) to improve the quality and usability of the above networks in the context of Smart Cities or Industrial scenarios.
Edge computing and software defined networking
The huge impact of Radio Access Network deployment on the CAPEX of infrastructure owners is being exacerbated in ultra-dense cellular networks. Moreover, in the aforementioned scenarios, the inefficiency of resources’ allocation has become one of the main drawbacks to boost the capacity of current and future cellular networks. The active Radio Access Network sharing based on Software-Defined Networking (SDN) emerges as a promising solution for 5G to cope with the ever-increasing infrastructure cost and the need for efficient resources allocation; however, proposals defined so far are preliminary and cannot guarantee the Quality of Experience (QoE) offered by the multiple tenants sharing the RAN.
The WINE group is focused on the design of algorithms, architectures and technical solutions to improve the efficiency in the use of network resources in multi-tenant scenarios, where the joint optimization of communication and processing resources has become a need. The research line works on the joint design of virtualization of networked processing elements (VNEs and VNRs), caching solutions, traffic slicing, mobile-edge computing, etc. We also focus on the edges of the 5G network including virtualization and software defined capillary networking elements to improve the operation of the last hop infrastructure.
Mobility and Radio Resources Management in 5G cellular networks
The main challenge of the future cellular networks (also known as 5G) is to meet the ever-increasing demand for massive connectivity and massive capacity. Although it has been shown that these two objectives must be addressed from a multi-fold approach, the densification of the network and the exploitation of new spectrum bands arise as two possible enablers.
In this context, the research community has clearly drawn the attention to the millimeter wave (mmWave) bands, where bandwidths of up to 1 GHz could be allocated to cellular systems. These new spectrum bands are essential to fuel the network capacity, but the propagation impairments at high frequencies, along with the highly directive antenna gains required to counteract them, pose new architectural and radio resources management challenges.
The group focuses, amongst others, on the design, development and evaluation of Medium Access Control (MAC) layers to provide efficient mechanisms in mmWave bands, e.g. cell discovery, cell association, self-backhauling, etc.