Title: Machine Learning for decentralised IOT networks
Abstract: The presentation will cover the research that has been done in IETR (Institute of Electronics and Telecommunications of Rennes, France, UMR CNRS 6164) on Cognitive Radio during the last 15 years, starting from very first Cognitive Radio concept, then deriving the Cognitive Cycle to be implemented in a radio device/system in order to make it support cognitive features. Then a special focus on Dynamic Spectrum Access (DSA) and especially Opportunitic Spectrum Access (OSA) will be done from the cognitive radio point of view, and on the learning and decision phase of the cognitive cycle point of view. We will show how we derived from our analysis ten years ago that OSA issue could be modeled as a Multi-Armed Bandit problem and how we validated that UCB (Upper Confidence Bound) algorithms could face radio constraints (including resistance to sensing false alarm and miss detection). OSA proof-of-concept done in the mid-2010s will be described. We will then analyze the difference between the OSA and IoT (Internet of Things) cognitive cycles in order to derive changes to be done for IoT. Proof-of-concept done early 2018 for IoT, based on USRP platforms, will be detailled, as well as a real implementation done in a LoRaWAN IoT network deployed at the University of Rennes at the end of year 2018. Our solution is a decentalized solution (learning is at device side) for spectrum sharing, without any protocol overhead (requiring no extra-transmissions, nor extra-data to be added in LoRaWAN frames) in IoT acknowledged mode.

Title: Cognitive Radar
Abstract: In the past few years, novel approaches to radar signal processing have been introduced which allow the radar signal detection and parameter estimation using much smaller number of measurements than required by Nyquist sampling. These systems exploit the fact that the target scene is sparse facilitating the use of recent advances in compressed sensing methods. This talk will introduce recent developments in reduced-rate sampling that break the link between common radar design trade-offs such as range resolution and transmit bandwidth; dwell time and Doppler resolution; spatial resolution and number of antenna elements; continuous-wave radar sweep time and range resolution. For each of these ideas, we present state-of-the-art hardware prototypes that we have designed and developed to demonstrate the real-time feasibility. We examine extensions to diverse applications such as cognition, spectral coexistence, matrix completion, autonomous driving, ground penetration radar, multiple-input-multiple-output and synthetic aperture radars. Finally, we connect the cognition in radars to some interesting applications of deep learning.

Title: Energy and spectrum efficiencies trade-off in wireless systems
Abstract: This talk regards the energy and spectral efficiencies trade-off in telecommunications. Starting with some fundamentals about the energy waste in the power amplifier, the benefits of CPM (Continuous Phase Modulation) systems (2B, Bluetooth, aeronautic telemetry, ...) are put into light. But the need for more spectral efficiencies justifies the use of OFDM is most of today's systems (DVB, 4G, 5G, WiFi). Consequently the peak to average power ratio (PAPR) of systems is presented as a key parameter to estimate their energy efficiencies. It is derived in many situations (single, multi-carriers, cognitive radio) and powerful methods to mitigate it are presented. The discussion ends by a theoretical study of NB-IoT systems in terms of energy efficiency analysis through the peak to average ratio (PAPR).

Title: Index Modulation for Next Generation Wireless Communication Systems
Abstract: This is an introductory talk on a scheme of information transfer called index modulation or IM in short. The IM has gained a recent attention and it is considered as a promising technology candidate for the next generation wireless communication systems. The basic idea underlying IM is to consider the subcomponents making up the physical layer of a communication system as forming a set. The transmitter of the IM scheme activates one, or in general a few, of these set members at a given transmission epoch using one, or in general several, bits to be transmitted. The receiver determines the indices of the members of the set that are activated at the transmitter and thereby demodulates the transmitted information bits. Two examples of the IM are Spatial Modulation or SM and Orthogonal Frequency Division Multiplexing with IM or OFDM-IM. The set members for the former are the transmit antennas, whereas they are the OFDM subcarriers for the latter. The talk provides an overview of SM and OFDM-IM. The talk also gives outline of a compressive sensing based approach for demodulating the transmitted information bits which is applicable for a generalized version of the SM.

Title: Interference and Energy Management Issues in Next Generation Wireless Systems
Abstract: With increasing push for higher spectrum utilization efficiencies, management of interference has become of utmost importance in next generation networks. We argue that the underlay framework provides a natural framework for interference management. Starting with a cellular setup in which two co-existing multiuser downlink networks transmit in the same frequency band as the cellular base-station, network management rules are derived for fixed and adaptive rate transmission with various degrees of channel knowledge. Insights into resource allocation are obtained. Another major challenge in evolving communication standards is lifetime of batteries in nodes. We show that use of energy harvesting to supplement battery energy along with careful battery energy management an maximise battery lifetime. Lastly, we analyze performance of self-sustaining nodes that have no regular battery, and can only store and use harvested energy. Using continuous-state Markov chains to model the energy in the buffer, insights are drawn into performance of communication links.

Title: Physical Layer Aware Link Layer Protocols
Abstract: Classical strategies at the link layer are oblivious to fading channel characteristics. While some recent approaches have addressed the need for physical layer awareness, truly dynamic cross-layering has not been fully exploited. In this presentation, we will first discuss how wireless channel is modeled for link layer communications. Then, we will walk through the link layer strategies that progressively look into the channel awareness, thereby achieving higher energy efficiency as well as channel utilization.

Title: Visible Light Communications - basics, importance, issues and solutions
Abstract: Visible light communications (VLC) is a viable supplement to existing radio frequency (RF) communications, since shifting communication towards the nm-wave range results in expansion of the available bandwidth to meet the spectrum and low power demands for the proposed 5G and beyond systems. In VLC, the light emitting diode (LED)- lamps, generally used as luminaries, additionally serve as optical-transmitters, and photodiode-arrays are used as optical receivers in a typical optical attocell. The intensity modulation of LED by users’ signals is done at a speed imperceptible to human eye, thereby achieving the dual goal of illumination and signal transmission at lower power as compared to RF equipment. VLC is a core component of Li-Fi systems, which has been proposed for applications like Internet of Things (IoT), 5G systems, underwater communications, vehicle-to-vehicle communication and many others. In this talk, we will go over the basics, importance, issues, and solutions for VLC, and also explore ways to increase the throughput with some future directions.

Title: IoT Based Automation Solutions for Distributed Infrastructure
Abstract: In this talk, I will give a brief overview about the IoT infrastructure that Zenatix has deployed for automating distributed chains like supermarkets, QSR and banks. I will then pick up some illustrative examples of how we derive actionable insights which can help the customers in driving energy savings and better asset management. I will then end with sharing some experiences around deploying and managing large scale IoT infrastructure in harsh conditions (such as intermittent network connectivity, poor power quality and power failures) in a country like India.

Title: 5G IOT Gateway
Abstract: LTE Cat 16 IoT Gateway Solution will enable a new world of connected applications with support of both sub-6GHz and mmWave spectrum. This advanced gateway provides LTE Cat 16 Internet connectivity via an 802.11n/ac /ax dual Wi-Fi network, providing users with Internet access with speeds up to 10 times faster than the current solutions.

Title: Energy Gathering in Practical Wireless Sensor Networks
Abstract: Minimizing energy consumption through efficient data gathering and energy harvesting are open research problems in practical wireless sensor networks (WSNs). This talk presents our recent work on integrated data and energy gathering framework for practical WSNs andis based on compressive sensing. In this, data is collected from a fewer sensor nodes randomly at any time point instead of picking up a linear combination of data from all sensor nodes at all time points. This reduces the computation cost, implementation complexity, energy losses at the sensor level while improving the data recovery. Also, a fewer sensors at any time point selected contribute to recovering data, while the rest of the sensor nodes harvest energy during those time points. This approach can be very helpful for practical WSNs that may be deployed in underwater or in forest areas where the nodes are starved of energy.

Title: Slice-Aware Millimeter-wave Selection for Sustaining the 5G Use-Cases
Abstract: Future wireless applications anticipate an explosion in the plethora of use-cases and services. Heterogeneous networks, comprising of multiple tiers of base stations (BSs) and operating in multiple radio access techniques (RATs) form an integral part of sustaining such diverse use-cases. One essential challenge of designing such complex networks is that of optimal user association to various tiers of BSs and to the available RATs. In the fourth generation (4G) networks, tier selection biasing is used mainly for load balancing. Offloading the user equipments (UEs) from the macro base station (MBS) to the small cell base stations (SBSs) is facilitated by a network-wide bias to expand the range of SBSs. On the contrary, in this talk, we discuss how to use various RAT-selection biases, each one associated to one of the fifth generation (5G) use-cases and designed to carefully satisfy the service requirements. In fact, to sustain the diverse use cases of 5G, a mobile operator will be able to define service-based logical partitions of its network over a common physical infrastructure. Network slicing facilitates the creation and management of such network instantiations (the network slices), each one composed by functions and parameters (e.g., the RAT bias in our work) tailored to address specific requirements [2]. In this talk, we focus on two specific RATs, i.e., sub-6GHz based LTE and mm-wave. We follow the specifications by 3GPP TS 23.501 [1], wherein, after the user, based on its service request, associates with one of the slices offered by the network, the bias value of the concerned slice is used for selection of the RAT in case the association is with an SBS. We will see how the diverse QoS requirements of different use-cases lead to a dramatically different RAT selectivity for the users.

Title: Are we ready yet for a smooth WiFi experience in busy settings? 802.11ax High Efficiency Next Generation WiFi
Abstract: In the last decade, the raw-bitrate of WiFi has seen a dramatic improvement from meager 1-2Mbps to few Gbps. But, the experience of using WiFi in dense settings such as stadium, railway station, busy airport is frustratingly poor. The new IEEE 802.11 standard i.e., 802.11ax aims to remedy exactly this precise problem. It proposes to improve the average performance of WiFi in dense settings by 4X by providing various key features such as multi-user transmission using OFDMA and MU-MIMO and spatial reuse.
In this talk, we will provide a brief overview of these key features and analyse how will they help in achieving the target.

Title: Global Navigation Satellite System: Principles and Applications
Abstract: The advent of Global Positioning System (GPS)\Global Navigation Satellite System (GNSS) technologies have radically changed the navigation process in a way that we take Positioning, Navigation and Timing solutions for granted, in many cases. This talk aims to introduce some concepts of the complex space-based navigation infrastructure, GPS\GNSS, GPS signal structure, acquisition and position computation techniques from GPS observations.
Since GPS is the oldest and most reliable among all available GNSS, the talk will start with an introduction to GPS. After discussing GPS signal structure various position error sources will be introduced and various error correction techniques will be explained. At the end of the talk, Navigation using Indian Constellation (NavIC) will be briefly discussed and several new research avenues will be explored.

Title: Age-of-Information: A Metric for Cyber Physical Systems
Abstract: In the talk we will start with an overview of the metric of age-of-information (AoI), its relevance to cyber physical system applications, and why it is distinct from commonly analyzed metrics of throughput and delay. We will discuss analysis of AoI for simple network abstractions followed by a brief look at end-to-end age control over the Internet. We will consider topical issues of coexistence of two networks that would like to optimize age and throughput respectively. The talk will end with the applications of sensing in large IoT networks and autonomous driving.

Title: Device to Device Communication: A new frontier for 5G cellular standards
Abstract: Device-to-device (D2D) is a new communication paradigm which has evolved recently to meet the capacity requirements of 4G/beyond 4G cellular networks. In this talk, I will discuss the spectrum sharing frameworks that were recently proposed to facilitate D2D communication in cellular networks. Specifically, this talk will focus on underlay, overlay and cooperative spectrum sharing frameworks, discuss the trade-off involved when selecting one framework over another and also deliberate on the coexistence issues involving D2D and cellular users. I will also share some of the current analytical, simulation and measurement results obtained by our research group in this direction.