Opendata, web and dolomites

Report

Teaser, summary, work performed and final results

Periodic Reporting for period 2 - TACTILENet (TACTILENet: Towards Agile, effiCient, auTonomous and massIvely LargE Network of things)

Teaser

We are witnessing an unprecedented worldwide growth of mobile data traffic that is expected to continue at an annual rate of 45% over the next years, reaching 30.5 exabytes per month by 2020. With the arrival of 5G, mobile connectivity goes from something we experience mainly...

Summary

We are witnessing an unprecedented worldwide growth of mobile data traffic that is expected to continue at an annual rate of 45% over the next years, reaching 30.5 exabytes per month by 2020. With the arrival of 5G, mobile connectivity goes from something we experience mainly through personal devices to being built into the fabric of our society, creating an integrated infrastructure that will connect buildings, transport, and utilities. While the demand for wireless capacity will continue to grow, the emergence of the Internet of Everything (IoE) system, connecting millions of people and billions of machines, is yielding a radical paradigm shift from the rate-centric enhanced mobile broadband (eMBB) services of the conventional networks towards ultra-reliable, low latency communications (URLLC). The objective of the TACTILENet project is to bring together the complementary expertise of European and third-country partners in order to lay the foundations for addressing basic issues in several facets of 5G and beyond networking.

Our conclusions of the action is summarized as below. We developed and analysed:
1. Edge-caching techniques, which bring the content closer to the users in order to improve their Quality of Experience.
2. Dynamic control techniques for networks with wireless energy transfer as well as energy-harvesting devices to provide perpetual and reliable connectivity for IoE.
3. Cross-layer techniques such as caching and massive-MIMO beamforming suitable for high bandwidth mobile communications.
4. System architectures involving airborne base stations and techniques for multiple-access to enable massive but short communications among machines.

Additionally, we have participated in standardisation bodies such as ITU, IEEE and Networld2020. We have partially completed a course on 5G networks.

Work performed

The researchers designed a coded caching architecture for heterogeneous networks incorporating finite sized caches at the edge nodes as well as user equipment. Unlike prior works, they considered random placement of content in the caches since this is more practically relevant, and identified the loss of optimality in this practical design. Finally, the researchers have considered the quality of experience for self-sufficient receivers when the receivers are powered by wireless energy transfer. Their work has demonstrated the fundamental limits of the point-to-point link capacity achieved by a dynamic control algorithm.

The researchers identified a new approach in scheduling user when users are harvesting their energy from sources that are correlated. The correlation among energy sources provides the information that can further facilitated to improve network throughput. The researchers have also investigated the effect of a correlated transmission channel when the transmitter is harvesting its energy from a renewable energy source. The researchers identified a simple index policy that controls transmission rate of the transmitter to significantly improve the network throughput. The researchers have investigated the control of network with simultaneous wireless power and information transfer under a reliability constraint. Transmitter employs Hybrid ARQ protocol for encoding of information where the receiver decides to use the incoming radio transmission for energy harvesting and for information decoding. The results indicate a simple to implement harvest-first-store-later type of policy minimizes the number of retransmissions in independent and identical channel conditions.

The researchers investigated the scenario where access nodes and/or number of communication links per unit area are densified with many Picocells / femtocells/ relays/ RRHs. In this scenario, Network function virtualization (NFV) is limited by the reliability of commercial off-the-shelf hardware. The researchers have proposed the idea of leveraging channel coding to enhance robustness on NFV to hardware failure. The researchers have demonstrated that the stationary channel models used for conventional MIMO systems are inadequate for large-aperture massive MIMO systems. Hence, they propose a simple visibility region based models to capture the non-stationarity in the channel, which shed light on the performance loss/gain due to non-stationarity. The researchers have developed different transceiver designs for massive MIMO with extremely large dimension. This work involved theoretical performance assessment as well as algorithmic development.

The researchers put forward a communication theoretic model in order to analyze the options for slicing of the radio resources and the access protocols. The researchers also study the reliability/latency trade-off in a coded slotted ALOHA system where the number of transmitting users is not known at the receiver. The optimal length and access probabilities are obtained by maximizing the expected probability of successful transmission for an arbitrary user. The researchers have collaboratively defined a novel concept of latency optimization in Virtual Reality (VR) environments. The researchers have developed and published one of the first works that treats random access algorithms with correlated activity of the users. This model is breaking the assumption adopted in random access algorithms since 1968, which states that the users in random access are activated in a random and uncorrelated manner. Our model with correlation starts to gain importance as the IoT devices start to proliferate, as their activation patterns are correlated.

Final results

It is well expected that tremendous increase in the number of 5G user equipment can only be supported by renewable energy sources to reduce the carbon footprint of telecommunications. The medium access control policies developed in this research have the potential to achieve this objective without significant loss of performance as compared to the conventional wireless communications systems.

Similarly, with the development of 5G communications systems, a colossal increase in mobile data is expected. Moreover, this mobile data needs of the users should be satisfied in ever shorter latencies. In fact, applications such as remote tactile operations will become reality. In this regard, caching of content in the network and ever closer to the users can help achieve this objective. To this end, the researchers of Tactilenet obtained the fundamental limits of cache-aided designs in heterogeneous wireless networks, which can be used to understand and compare the performances of practical algorithms.

The socio-economic impact and the wider societal implications of the project include reduced energy consumption by Telecom Operators due to increased efficiency of their operations as well as the harvesting of natural energy resources thanks to the protocols developed within the project; improved reliability and latency of communications which has the potential to change how we feel and understand our environment by introducing for the first time relaying tactile signals from distance; faster and ubiquitous access to information by a wide range of citizens located in both rural and metropolitan areas.

Website & more info

More info: http://tactilenet.sabanciuniv.edu.