An NB-IoT-Based Edge-of-Things Framework for Energy-Efficient Image Transfer.

Machine Learning (ML) techniques can play a pivotal role in energy efficient IoT networks by reducing the unnecessary data from transmission. With such an aim, this work combines a low-power, yet computationally capable processing unit, with an NB-IoT radio into a smart gateway that can run ML algorithms to smart transmit visual data over the NB-IoT network. The proposed smart gateway utilizes supervised and unsupervised ML algorithms to optimize the visual data in terms of their size and quality before being transmitted over the air. This relaxes the channel occupancy from an individual NB-IoT radio, reduces its energy consumption and also minimizes the transmission time of data. Our on-field results indicate up to 93% reductions in the number of NB-IoT radio transmissions, up to 90.5% reductions in the NB-IoT radio energy consumption and up to 90% reductions in the data transmission time.

Optical Detection Methods for High-Throughput Fluorescent Droplet Microflow Cytometry.

High-throughput microflow cytometry has become a focal point of research in recent years. In particular, droplet microflow cytometry (DMFC) enables the analysis of cells reacting to different stimuli in chemical isolation due to each droplet acting as an isolated microreactor. Furthermore, at high flow rates, the droplets allow massive parallelization, further increasing the throughput of droplets. However, this novel methodology poses unique challenges related to commonly used fluorometry and fluorescent microscopy techniques. We review the optical sensor technology and light sources applicable to DMFC, as well as analyze the challenges and advantages of each option, primarily focusing on electronics. An analysis of low-cost and/or sufficiently compact systems that can be incorporated into portable devices is also presented.

Q-Learning Based Joint Energy-Spectral Efficiency Optimization in Multi-Hop Device-to-Device Communication.

In scenarios, like critical public safety communication networks, On-Scene Available (OSA) user equipment (UE) may be only partially connected with the network infrastructure, e.g., due to physical damages or on-purpose deactivation by the authorities. In this work, we consider multi-hop Device-to-Device (D2D) communication in a hybrid infrastructure where OSA UEs connect to each other in a seamless manner in order to disseminate critical information to a deployed command center. The challenge that we address is to simultaneously keep the OSA UEs alive as long as possible and send the critical information to a final destination (e.g., a command center) as rapidly as possible, while considering the heterogeneous characteristics of the OSA UEs. We propose a dynamic adaptation approach based on machine learning to improve a joint energy-spectral efficiency (ESE). We apply a Q-learning scheme in a hybrid fashion (partially distributed and centralized) in learner agents (distributed OSA UEs) and scheduler agents (remote radio heads or RRHs), for which the next hop selection and RRH selection algorithms are proposed. Our simulation results show that the proposed dynamic adaptation approach outperforms the baseline system by approximately 67% in terms of joint energy-spectral efficiency, wherein the energy efficiency of the OSA UEs benefit from a gain of approximately 30%. Finally, the results show also that our proposed framework with C-RAN reduces latency by approximately 50% w.r.t. the baseline.

Adaptive LINE-P: An Adaptive Linear Energy Prediction Model for Wireless Sensor Network Nodes.

In the context of wireless sensor networks, energy prediction models are increasingly useful tools that can facilitate the power management of the wireless sensor network (WSN) nodes. However, most of the existing models suffer from the so-called fixed weighting parameter, which limits their applicability when it comes to, e.g., solar energy harvesters with varying characteristics. Thus, in this article we propose the Adaptive LINE-P (all cases) model that calculates adaptive weighting parameters based on the stored energy profiles. Furthermore, we also present a profile compression method to reduce the memory requirements. To determine the performance of our proposed model, we have used real data for the solar and wind energy profiles. The simulation results show that our model achieves 90-94% accuracy and that the compressed method reduces memory overheads by 50% as compared to state-of-the-art models.

Dual-Source Linear Energy Prediction (LINE-P) Model in the Context of WSNs.

Energy harvesting technologies such as miniature power solar panels and micro wind turbines are increasingly used to help power wireless sensor network nodes. However, a major drawback of energy harvesting is its varying and intermittent characteristic, which can negatively affect the quality of service. This calls for careful design and operation of the nodes, possibly by means of, e.g., dynamic duty cycling and/or dynamic frequency and voltage scaling. In this context, various energy prediction models have been proposed in the literature; however, they are typically compute-intensive or only suitable for a single type of energy source. In this paper, we propose Linear Energy Prediction "LINE-P", a lightweight, yet relatively accurate model based on approximation and sampling theory; LINE-P is suitable for dual-source energy harvesting. Simulations and comparisons against existing similar models have been conducted with low and medium resolutions (i.e., 60 and 22 min intervals/24 h) for the solar energy source (low variations) and with high resolutions (15 min intervals/24 h) for the wind energy source. The results show that the accuracy of the solar-based and wind-based predictions is up to approximately 98% and 96%, respectively, while requiring a lower complexity and memory than the other models. For the cases where LINE-P's accuracy is lower than that of other approaches, it still has the advantage of lower computing requirements, making it more suitable for embedded implementation, e.g., in wireless sensor network coordinator nodes or gateways.

A QoS Optimization Approach in Cognitive Body Area Networks for Healthcare Applications.

Wireless body area networks are increasingly featuring cognitive capabilities. This work deals with the emerging concept of cognitive body area networks. In particular, the paper addresses two important issues, namely spectrum sharing and interferences. We propose methods for channel and power allocation. The former builds upon a reinforcement learning mechanism, whereas the latter is based on convex optimization. Furthermore, we also propose a mathematical channel model for off-body communication links in line with the IEEE 802.15.6 standard. Simulation results for a nursing home scenario show that the proposed approach yields the best performance in terms of throughput and QoS for dynamic environments. For example, in a highly demanding scenario our approach can provide throughput up to 7 Mbps, while giving an average of 97.2% of time QoS satisfaction in terms of throughput. Simulation results also show that the power optimization algorithm enables reducing transmission power by approximately 4.5 dBm, thereby sensibly and significantly reducing interference.