Health and sociological indicators alert that life expectancy is increasing, hence so are the years that patients have to live with chronic diseases and co-morbidities. With the advancement in ICT, new tools and paradigms are been explored to provide effective and efficient health care. Telemedicine and health sensors stand as indispensable tools for promoting patient engagement, self-management of diseases and assist doctors to remotely follow up patients. In this paper, we evaluate a rapid prototyping solution for information merging based on five health sensors and two low-cost ubiquitous computing components: Arduino and Raspberry Pi. Our study, which is entirely described with the purpose of reproducibility, aimed to evaluate the extent to which portable technologies are capable of integrating wearable sensors by comparing two deployment scenarios: Raspberry Pi 3 and Personal Computer. The integration is implemented using a choreography engine to transmit data from sensors to a display unit using web services and a simple communication protocol with two modes of data retrieval. Performance of the two set-ups is compared by means of the latency in the wearable data transmission and data loss. PC has a delay of 0.051 ± 0.0035 s (max = 0.2504 s), whereas the Raspberry Pi yields a delay of 0.0175 ± 0.149 s (max = 0.294 s) for N = 300. Our analysis confirms that portable devices ( $p<<0.01$ ) are suitable to support the transmission and analysis of biometric signals into scalable telemedicine systems.

With the increasingly dense-deployed small cells, the heterogeneous ultra-dense networks (H-UDN) now faces with new challenges. In order to fulfill user traffic demand and improve network energy efficiency in H-UDNs, this paper researches on the adaptive cell zooming scheme to achieve the optimized user association through adjusting cell coverages. Then, the cell sleeping scheme is further applied to turn off light traffic load cells for Base Station (BS) power saving. We define the Cell Zooming Factor (CZF) to express cell coverage, which is derived as a closed-form expression with the relationship of small cell and macro cell densities. The user association probability with one typical cell is also derived, where the probability is related with both the transmission power and CZF. We apply the game theory to optimize the CZF based on reduction of area power consumption. Based on the fact that users are encouraged to associate with small cells to reduce transmission power consumptions, the small cell sleeping scheme is further applied to improve the network energy efficiency, while the sleeping small cell users could be offloaded to macro cells. The sleeping probability is defined and derived with influences of different sleeping thresholds. We optimize the threshold with maximized energy saving performance. Simulation results verify that the sleeping threshold has an optimized value.

The electric vehicle (EV) industry has been rapidly developing internationally due to a confluence of factors, such as government support, industry shifts, and private consumer demand. Envisioning for the future connected vehicles, the popularity of EVs will have to handle a massive information exchange for charging demand. This inevitably brings much concern on network traffic overhead, information processing, security, etc. Data analytics could enable the move from Internet of EVs to optimized EV charging in smart transportation. In this paper, a mobile edge computing (MEC) supporting architecture along with an intelligent EV charging recommendation strategy is designed. The global controller behaves as a centralized cloud server to facilitate analytics from charging stations (CSs) (service providers) and charging reservation of on-the-move EVs (mobile clients) to predict the charging availability of CSs. Besides, road side units behave as MEC servers to help with the dissemination of the CSs’ charging availability to EVs, and collecting their charging reservations, as well as operating decentralized computing on reservations mining and aggregation. Evaluation results show the features of the MEC-based charging recommendation system in terms of communication efficiency (low cost for information dissemination and collection) and improvement of charging performance (reduced charging waiting time and increased fully charged EVs).

As a typical Internet-of-Things (IoT) scenario, massive machine-type communications (mMTC) services are expected to grow exponentially and create a multi-billion dollar industry spanning a broad range of vertical sectors. In literature, wireless software-defined-network (SDN) is viewed as a promising approach to facilitate the degree of reconfigurability on extended sets of mMTC devices via centralized software updates. However, most of the current anomaly detection scheme (ADS) in SDN suffers from the high risk of overwhelming of the controller as well as excessive energy consumption if directly applied in the network with enormous devices. To address the scalability issues in centralized ADS, we propose a localized ADS scheme, called scalable and energy-efficient anomaly detection scheme (SEE-ADS), comprising of a detection activation module, a lightweight pre-detection module, a heavyweight anomaly detection module, a dynamic strategy selection module. Through the cooperation among these modules, the proposed ADS is capable of detecting attacks dynamically and effectively without the risk of energy depletion via discontinuous activation of the heavyweight detection. The lower complexity is fulfilled by developing a localized and adaptive heavyweight detection module, called Localized Evolving Semi-supervised Learning Based Anomaly detection scheme (LESLA). Besides, the proposed scheme makes full use of feedback from the previous heavyweight activation and the indication of pre-detection on each packet. Simulation results show that the proposed scheme greatly reduces the overall energy consumption over heavyweight detection. Furthermore, the proposed scheme shows higher sensitivity on abnormal packets and similar false alarm compared with the literature work.

In the era of IoT, reinforcement learning (RL) based techniques are promising candidates to handle the intrusion response through the interaction between the IoT device and its environment. Given a large number of devices in IoT network, wireless software-defined-networking (W-SDN) is widely agreed to be introduced to facilitate the network management, such as launching a new intrusion response scheme (IRS) on massive IoT devices. To guarantee the scalability and security of IoT services on the extended devices in W-SDN-based network, a distributed RL-based IRS is proposed in this paper, called game-theoretic actor-critic-based IRS (GTAC-IRS). GTAC-IRS employs a game-theoretic based response selection matrix, aiming at reducing training time and facilitating the convergence of the response scheme. GTAC-IRS constructs a well-designed state representation of observed environment status, a low dimension response matrix and a simplified response selection policy to lower the complexity of the algorithms. Simulation results reveal that benefiting from local environment observation, GTAC-IRS achieves effective intrusion response without sophisticated feature engineering. Instead of the ``warm-startb'' training adopted in conventional RL-based IRS, the low dimension response matrix in GTAC-IRS can significantly improve the convergence speed. Thus, GTAC-IRS outperforms other popular IRSs in terms of the response time under the circumstance of the node’s behaviour changing or malicious nodes ratio changing.