Search

Recently, the advances in passive brain-computer interfaces (BCIs) based on electroencephalogram (EEG) have shed light on real-world neuromonitoring technologies. However, human variability in the EEG activities hinders the development of practical applications of EEG-based BCI. To tackle this problem, many transfer-learning techniques perform supervised calibration. This kind of calibration approach requires task-relevant data, which is impractical in real-life scenarios such as drowsiness during driving. This study presents a transfer-learning framework for EEG decoding based on the low-dimensional representations of subjects learned from the pre-trial EEG. Tensor decomposition was applied to the pre-trial EEG of subjects to extract the underlying characteristics in subject, spatial, and spectral domains. Then, the proposed framework assessed the characteristics to obtain the low-dimensional subject representations such that the subjects with similar brain dynamics can be identified. This method can leverage the existing data from other users, and a small number of data from a rapid, non-task, unsupervised calibration from a new user to build an accurate BCI. Our results demonstrated that, in terms of prediction accuracy, the proposed low-dimensional subject representation-based transfer learning (LDSR-TL) framework outperformed the random selection, and the Riemannian manifold approach in cognitive-state tracking, while requiring fewer training data. The results can greatly improve the practicability, and usability of EEG-based BCI in the real world.

Extending the two-dimensional topology of wireless networks to the three-dimensional (3D) structure has recently been recognized as a revolutionary breakthrough for the next generation of wireless communication networks. As revealed in the literature, dispatching the unmanned aerial vehicle (UAV) assisted base-station (UBS) can fill in the coverage holes as the low energy-efficiency (EE) BSs on the ground are switched off. De facto, properly deploying the UBSs can further enlarge the flexibility of the user handover and exchange procedures in the sleep strategy. Thus, in this paper, we propose a low-complexity UBS-assisted sleep strategy for the small-cell network (SCN). Verified by the simulation results, it is found that deploying a proper number of capable UBSs can remarkably raise the EE for the 3D SCNs.

In the recent years, we have witnessed an explosion of networking applications due to the reasons such as the rapid development of cloud infrastructure, edge computing, and the Internet of Things. Furthermore, those applications become complex, the problem related to the large size of the state space and limited metric collection has emerged. This leads to an urging demand for adaptive management method in network routing. However, the complexity of traditional routing algorithms can be prohibited for practical systems. To overcome this challenge, we propose a novel tensor-based reinforcement learning method to route and schedule the packet flows, which is adaptive and model-free. Moreover, we improve the learning quality and efficiency by combining the Tucker decomposition technique within the learning process so that the machine learning direction can be obtained with low complexity. Finally, simulation results show that our proposed algorithm can achieve better performance under the same training episode and more stable results with less convergence time than conventional routing method, K-shortest path, traditional reinforcement learning approaches (i.e. Q-learning and SARSA) and comparable results to DQL.

Ultra Reliable and Low Latency Communications (URLLC) play a key role in 5G vertical markets, but pose many technical challenges especially when sharing the spectrum with Enhanced Mobile Broadband (eMBB) customers. This study aims to overcome the spectrum inefficiency issue of fully separate (FS) approach and the contention issue of the fully overlap (FO) approach. We present a user-initiated probability elastic resource (UPER) approach by dynamically adjusting the probability of using the shared spectrum for eMBB and URLLC traffic based on the current success and failure status of packet transmission status. The probabilities of successful transmission are derived for UPER, FS, and FO and partially overlap (PO) sharing spectrum approaches. We find that the successful transmission probability of UPER approach is 28% and 46% higher than FS and FO approaches, respectively. We further evaluate the reliability and throughput performance of URLLC and eMBB. When the URLLC packet load is low, the UPER method can almost achieve the best performance of the FS method. When the URLLC packet load is high, we show that UPER can improve the reliability performance up to 54% compared with other methods.

Latent tree model (LTM) is a probabilistic tree-structured graphical model, which can reveal the hidden hierarchical causal relations among data contents and play a key role in explainable artificial intelligence. However, because current LTM modeling techniques are only suitable for single-content variable, the applications of LTMs are somewhat limited. Toward this end, a multilayer LTM (ML-LTM) is first presented to deal with the hierarchical clustering issues of multicontent variables. Second, we further develop an ML-LTM-based multicontent recommendation system. Our experiment results show that the proposed ML-LTM can achieve 90% recommendation accuracy, but the current LTM can only has 20%. Third, we propose an incremental update approach for ML-LTM that can save five-sixth updating time comparing with the whole-model retraining approach for achieving the same recommendation accuracy.

The halogen-free flame retardant (HFFR) thermoplastic elastomers (TPEs) composites filled with aluminum hypophosphite had been prepared by twin-screw extruder. The flame retardancy, thermal stability, charring-effect, mechanical and rheological properties of TPE composites had been investigated by cone calorimeter, limiting oxygen index (LOI), UL-94 (vertical flame) test, thermogravimetric analysis (TGA), scanning electron microscopy (SEM), ultimate tensile and rheometer. The research results indicated that the flame retardancy and thermal stability of flame retardant (FR) TPEs had been dramatically improved. Obvious char-forming effect can be found for the samples with the addition of polyphenylene oxide (PPO) during high heat flow conditions, which presented more content and more compact of residue compared to the samples without PPO after combustion of cone calorimeter. Mechanical and rheological characteristics showed that the tensile strength and processability can be well maintained for FR TPEs composites, which was very important for industrial applications.

Millimeter wave (mmWave) technology achieving multi-gigabits speed plays a significant role in beyond 5G and the next 6G wireless communication networks thanks to its huge spectrum utilization and beam-based directional transmissions. To tackle temporary ultra-high data demands of hotspot areas, three-dimensional (3D) heterogeneous network (HetNet) is designed with the integration of mmWave unmanned aerial vehicles (UAV) to provide resilient instantaneous control and data transmissions. However, some critical beam-related issues for mmWave implementation of UAVs/drones are addressed including robust initial beam alignment, mobility-aware beam tracking and beam refinement. In this research, we aim at developing robust and efficient beam control mechanisms by implementing a prototype of 3D flying heterogeneous communications. The backhaul connections operate at mmWave frequency between airship and UAV/drone, while fronthaul links adopt lower frequency bands such as Wi-Fi for multiuser data transmissions. We evaluate system performances for our proposed beam control schemes and provide a real-time prototype of 3D on-demand flying mobile communication for mmWave HetNets.

Drone small cells (DSCs) can provide on-demand air-to-ground wireless communications in various unexpected situations, such as traffic jam or natural disasters. However, a DSC needs to face the challenges such as severe co-channel interference, limited battery capacity, and fast topology changes. Aiming to improve energy efficiency of DSCs and quality of services of customers, this paper presents a learning-based multiple drone management (LDM) framework by controlling the transmission power and the 3-dimension location of DSCs based on location data, and reference signal received power of users. Since the labeled throughput data are typically not available in emergency situations, we develop unsupervised learning DSC management techniques: 1) affinity propagation interference management scheme to mitigate interference and energy consumption, and 2) K-means position adjustment to adjust the new 3-dimension positions of drones. Our numerical results show that the proposed LDM framework combining with affinity propagation clustering and k-means clustering can enhance the energy efficiency of DSCs by 25% and the signal-to-interference-plus-noise ratio of ground users by 56%, respectively.

The dense deployment of small cells in 5G cellular networks raises the issue of controlling downlink inter-cell interference under time-varying channel states. In this paper, we propose a reinforcement learning based power control scheme to suppress downlink inter-cell interference and save energy for ultra-dense small cells. This scheme enables base stations to schedule the downlink transmit power without knowing the interference distribution and the channel states of the neighboring small cells. A deep reinforcement learning based interference control algorithm is designed to further accelerate learning for ultra-dense small cells with a large number of active users. Analytical convergence performance bounds including throughput, energy consumption, inter-cell interference, and the utility of base stations are provided and the computational complexity of our proposed scheme is discussed. Simulation results show that this scheme optimizes the downlink interference control performance after sufficient power control instances and significantly increases the network throughput with less energy consumption compared with a benchmark scheme.

Wireless and mobile technologies, as well as their users, are growing rapidly as the Internet of Things (IoT) products, such as sensor-network technologies, mobile devices, and supporting applications, become widely dispersed. Owing to the dynamic changes in the wireless networks and the exponential growth of the IoT products, which make it difficult to locate large quantities of users and devices, providing accurate tracking and trajectory predictions in open and highly condensed wireless networks is extremely difficult. An adaptive and scalable system is required to offer accurate location-based services (LBSs) for the success of IoT. To enhance the attainment of IoT, we propose a hybrid of principal component analysis (PCA) and gated recurrent unit (GRU) algorithms for mobility predictions in a wireless urban area. During the system development processes, we first collect an LTE signal from three unmanned aerial vehicle base stations (UAV-BSs), the Wi-Fi signal strength from each reachable Wi-Fi access points (APs), and channel information from the Wi-Fi signal media. We then apply PCA to reduce the number of Wi-Fi features and to decrease signal noise. Next, we train the GRU algorithm to develop models that can predict the mobility of IoT device users. Finally, we evaluate the tracking and trajectory models. To evaluate the proposed techniques, we compare the common parameters of the GRU with those of other deep learning types. The proposed technique provides plausible and state-of-the-art results for mobility predictions of IoT devices in a wireless environment.

This paper proposes a blockchain-based data checking scheme to protect data integrity in Internet of Things (IoT). Traditional data integrity schemes such as symmetric key approaches and public key infrastructure (PKI) suffer from the single-point of failure and network congestion due to the centralized architecture. Motivated by the distributed data authentication in blockchain, we propose to adopt blockchain to ensure the data integrity in IoT networks. However, the existing blockchain scheme cannot be directly applied to IoT nodes with limited computing and network resources. Hence, we develop a stochastic blockchain scheme to limit the number of cooperative nodes and distribute the load to IoT edge nodes. In our scheme, the IoT data are broadcast by randomly selected cooperative nodes, thereby introducing uncertainty to the attacker and improving the system security level. Finally, we propose a lightweight mining process to make only the IoT edge nodes compete for block generation and share the block with other nodes. If our scheme is used in the case of having 9,000 legitimate nodes and 1,000 compromised nodes, only three cooperative nodes can achieve the probability of successful defense over 99 percent.

Editor’s note: Multiband radio-frequency interconnect (MRFI) is an emerging technology to achieve low-latency and energy-efficient on-chip communication. This article proposes a BIST method to improve the reliability of MRFI-based design. – Partha Pratim Pande, Washington State University

In this letter, we consider the concept of mobile crowd-machine learning (MCML) for a federated learning model. The MCML enables mobile devices in a mobile network to collaboratively train neural network models required by a server while keeping data on the mobile devices. The MCML thus addresses data privacy issues of traditional machine learning. However, the mobile devices are constrained by energy, CPU, and wireless bandwidth. Thus, to minimize the energy consumption, training time, and communication cost, the server needs to determine proper amounts of data and energy that the mobile devices use for training. However, under the dynamics and uncertainty of the mobile environment, it is challenging for the server to determine the optimal decisions on mobile device resource management. In this letter, we propose to adopt a deep $Q$ -learning algorithm that allows the server to learn and find optimal decisions without any a priori knowledge of network dynamics. Simulation results show that the proposed algorithm outperforms the static algorithms in terms of energy consumption and training latency.

With the rapid growth of Internet of Things (IoT), integrating a variety of IoT can result in novel applications. However, IoT devices are often deployed in an open environment where IoT are inclined to be malfunctioned. Although data reliability can be achieved by data recovery with conventional replication, the communication between IoT is susceptible to eavesdropping. Therefore, in this paper, we study the eavesdropping prevention of data repair in IoT environments based on network coding. We theoretically derive the relation between security level and storage in heterogeneous IoT systems. To further reduce the repair bandwidth, we exploit recurrent neural network for the storage failure prediction. Under the condition when failure probability and workloads of storage devices are considered, two allocation algorithms are proposed to avoid data repair. Finally, we show the relation between storage cost and reliability with different numbers of IoT devices. Experimental results manifest that the proposed allocation algorithms can outperform the baseline case by 18.4% in terms of the security level.

Deploying dense small cells is the key to providing high capacity, but raise the serious issue of energy consumption and inter-cell interference. To understand the behaviors of ultra-dense small cells (UDSC) with dynamic interference and traffic patterns, this paper presents a data-driven resource management (DDRM) framework to implement power control and channel rearrangement in UDSC. We find that the inter-cell interference can be used to describe the affinity of cells. Thus, we propose an unsupervised learning algorithm for UDSC, called affinity propagation power control (APPC) mechanism. In principle, APPC first groups small cells into different clusters and identifies cluster centers. Next, the transmission power of a cluster center is decreased to reduce the interference to the neighboring cells’ users in this cluster. Since lowering transmission power of a cluster center cell may cause the performance degradation to the users at the cell edge, a victim-aware channel rearrangement (VACR) mechanism is further designed to adjust the channel usage bandwidth of the neighboring cells in order to guarantee the quality of service of these victimized users. Our simulation results show that the DDRM framework can significantly improve energy efficiency and throughput in UDSC compared to the existing approaches.

This paper presents a data-driven biadaptive self-organizing network (Bi-SON) for ultra-dense small cells (UDSC), which can improve energy efficiency and reduce interference in dynamic environments, taking account of cell switching on / off , transmission power adjustment, and traffic loads simultaneously. In the first adaptation of Bi-SON, a joint traffic load and interference aware cell ranking mechanism first determines the necessary active small cells based on traffic loads, and then ranks all the active small cells based on their carried traffic load and resulting interference. Top ranked cells will transmit at the maximum power. The last ranked $K$ cells will adjust the transmission power for interference reduction in the second adaptation function of Bi-SON, while maintaining the required quality of service. According to a polynomial regression learning approach, the total system throughput of UDSC is characterized as a function of $K$ . Compared to the baseline case when all the cells transmit with the maximum power, our proposed Bi-SON framework can improve the throughput and energy efficiency of UDSC by 73% and 169%, respectively. However, the pure switching on/off approach can only improve the throughput and the energy efficiency of UDSC by 52% and 115%, respectively. As demonstrated, even with a simple power adaptation algorithm, a learning-based Bi-SON framework can improve the performance of UDSC by taking advantage of the pervasive availability of voluminous data.

Software defined networking (SDN) enables network function programmability for ease of configuration and maintenance, and also allows network administrators to change traffic rules on the fly. However, denial of service (DoS) attacks pose security challenges on the centralized control plane of SDN. Although the transport layer security (TLS) can help secure the control plane, it is computationally intensive, complex to configure, and not mandatory in OpenFlow protocol. In this paper, we present a lightweight authentication solution, called hidden authentication (HiAuth), to protect the SDN controller by hiding the identities of the forwarding devices into the control packets via efficient bitwise operations. HiAuth is the first to incorporate information hiding techniques into OpenFlow to provide security against DoS attacks. HiAuth exploits the IP identification field of IPv4 and the transaction identification field of OpenFlow in two authentication schemes. The experimental results show that HiAuth can effectively mitigate intruder DoS attacks and provide high undetectability to attackers.

Future-generation wireless networks (5G and beyond) must accommodate surging growth in mobile data traffic and support an increasingly high density of mobile users involving a variety of services and applications. Meanwhile, the networks become increasingly dense, heterogeneous, decentralized, and ad hoc in nature, and they encompass numerous and diverse network entities. Consequently, different objectives, such as high throughput and low latency, need to be achieved in terms of service, and resource allocation must be designed and optimized accordingly. However, considering the dynamics and uncertainty that inherently exist in wireless network environments, conventional approaches for service and resource management that require complete and perfect knowledge of the systems are inefficient or even inapplicable. Inspired by the success of machine learning in solving complicated control and decision-making problems, in this article we focus on deep reinforcement- learning (DRL)-based approaches that allow network entities to learn and build knowledge about the networks and thus make optimal decisions locally and independently. We first overview fundamental concepts of DRL and then review related works that use DRL to address various issues in 5G networks. Finally, we present an application of DRL for 5G network slicing optimization. The numerical results demonstrate that the proposed approach achieves superior performance compared with baseline solutions. Accepted version

This paper proposes an enhanced secure pseudonym scheme to protect the privacy of cloud data in Internet of Drones (IoD). Nowadays, drones equipped with cameras can provide surveillance and aerial photography applications. Unlike the video devices, personal drones with high mobility can track and follow an individual, causing both the identity and location privacy issues. However, IoD devices cannot implement complex cryptographic schemes because of limited computing power. To this end, we develop a secure light-weight network coding pseudonym scheme. Our designed two-tier network coding can decouple the stored IoD cloud data from the owner’s pseudonyms. Therefore, our proposed network coding-based pseudonym scheme can simultaneously defend against both outside and inside attackers. We implement our proposed two-tier light-weight network coding mechanism when facing untrusted cloud database. Compared to the computationally secure hash-based pseudonym scheme, our proposed scheme achieves the highest unconditional security level, but also can reduce more than 90% of processing time as well as 10% of energy consumption.

With the rapid advances in machine learning algorithms and sensing technologies, machine prognostics and health management (PHM) via data-driven approaches has become a trend in sophisticated machi...

In this paper, we present a topology-aware network coding (TANC) technique for wireless multicast systems aiming at reducing packet retransmissions in a lossy propagation environment. In addition to radio channel impacts, packet loss may result from traffic congestion, thereby yielding different retransmission requests. In such a wireless network coding multicast environment, what is the best packet combination to minimize the packet retransmissions for a number of users? We make progress on this problem by taking advantage of the network topology information to select the appropriate lost encoded packets in the encoding phase. We leverage the capability of software-defined networking controller to construct the global network topology in the proposed TANC scheme. Experimental results show that the proposed TANC scheme can reduce 52.4% of packet retransmissions compared to the existing random linear network coding in different environments, including various numbers of uncoded packets, network nodes, and packet loss probabilities.

In the hospital, a sleep postures monitoring system is usually adopted to transform sensing signals into sleep behaviors. However, a home-care sleep posture monitoring system needs to be user friendly. In this paper, we present iSleePost&mdash, a user-friendly home-care intelligent sleep posture monitoring system. We address the labor-intensive labeling issue of traditional machine learning approaches in the training phase. Our proposed mobile health (mHealth) system leverages the communications and computation capabilities of mobile phones for provisioning a continuous sleep posture monitoring service. Our experiments show that iSleePost can achieve up to 85 percent accuracy in recognizing sleep postures. More importantly, iSleePost demonstrates that an easy-to-wear wrist sensor can accurately quantify sleep postures after our designed training phase. It is our hope that the design concept of iSleePost can shed some lights on quantifying human sleep postures in the future.

Solar-powered and power grid connected green cellular networks are becoming attractive due to low carbon footprint and cost-effectiveness in providing uninterrupted service. In this paper, we analyze the performance of such dualpowered multi-cell network in presence of skewed traffic load across the different base stations (BSs). Cell coverage is decided at the network design stage based on long-term average traffic intensity across the various regions of a multi-cell coverage area. In presence of dynamically-changing skewness of traffic loads across different cells, we propose to adjust the cell coverage to accommodate the traffic and energy availability imbalance in the cells, while the demand for residual energy deficiency for serving the customers is fulfilled through the power grid connectivity. Network service provider’s cost with the proposed coverage adjustment based strategy is compared with that of the conventional approach where the individual BSs do not undergo any cell coverage adjustment and seek to provide the maximum network performance. Our analysis and simulation-based performance results demonstrate, that the network performance as well as monetary gains of the service provider are significantly higher with our proposed strategy. For example, at a moderate (30%) traffic skewness, the proposed strategy offers about 4% gain in operator’s annual profit, while serving about 8% more users on average at the peak hour. At a very high (80%) skewness, these numbers are respectively about 50% and 39%.

With the advancement of technology, the data generated in our lives is getting faster and faster, and the amount of data that various applications need to process becomes extremely huge. Therefore, we need to put more effort into analyzing data and extracting valuable information. Cloud computing used to be a good technology to solve a large number of data analysis problems. However, in the era of the popularity of the Internet of Things (IoT), transmitting sensing data back to the cloud for centralized data analysis will consume a lot of wireless communication and network transmission costs. To solve the above problems, edge computing has become a promising solution. In this paper, we propose a new algorithm for processing probabilistic skyline queries over uncertain data streams in an edge computing environment. We use the concept of a second skyline set to filter data that is unlikely to be the result of the skyline. Besides, the edge server only sends the information needed to update the global analysis results on the cloud server, which will greatly reduce the amount of data transmitted over the network. The results show that our proposed method not only reduces the response time by more than 50% compared with the brute force method on two-dimensional data but also maintains the leading processing speed on high-dimensional data., 6 pages, 5 figures, to appear in 2020 IEEE Global Communications Conference: Selected Areas in Communications: Big Data

This paper has been presented at 2020 European Conference on Networks and Communications (EuCNC). The fifth generation (5G) mobile network has paved the way for innovations across vertical industries. The integration of distributed intelligent edge into the 5G orchestrated architecture brings the benefits of low-latency and automation. A successful example of this integration is exhibited by the 5G-DIVE project, which aims at proving the technical merits and business value proposition of vertical industries such as autonomous drone surveillance and navigation. In this paper, and as part of 5G-DIVE, we present an aerial disaster relief system, called EagleEYE, which utilizes edge computing and machine learning to detect emergency situations in real-time. EagleEYE reduces training time by devising an object fusion mechanism which enables reusing existing datasets. Furthermore, EagleEYE parallelizes the detection tasks to enable real-time response. Finally, EagleEYE is evaluated in a real-world testbed and the results show that EagleEYE can reduce the inference latency by 90% with a high detection accuracy of 87%. This work has been partially funded by the H2020 EU/TW joint action 5G-DIVE (Grant #859881).

Recently, the development of Unmanned Aerial Vehicle (UAV) has been nearly matured and widely used in various fields. The combination of UAV and communication technologies, such as UAV Base Station (UAV-BS), can significantly increase the flexibility and scalability of the overall communication networks to provide more efficient communication services. While the UAV-BS improves the network service efficiency, the quality of services (QoS) in the air-to-ground communication link is highly affected unless the right users are unknown. In this paper, we propose the learning-based downlink user selection algorithm. The 3D downlink channel can be fast identified to judiciously select the users subset. In our proposed framework, we combine the k-means clustering and Convolutional Neural Network (CNN) that can increase the estimation accuracy of 3D wireless channels to enhance the communication service efficiency of the UAV-BS network. The field measurement results show that proposed method can achieve an average bit error rate (BER) of 3.56x10−7, which is better than the distance-based selection scheme that has an average of BER 2.88x10−3. The feasibility and effectiveness of the proposed method in real environment are proved, experimentally.

Future mobile communication networks require an Aerial Base Station (ABS) with fast mobility and long-term hovering capabilities. At present, unmanned aerial vehicles (UAV) or drones do not have long flight times and are mainly used for monitoring, surveillance, and image post-processing. On the other hand, the traditional airship is too large and not easy to take off and land. Therefore, we propose to develop an "Artificial Intelligence (AI) Drone-Cruiser" base station that can help 5G mobile communication systems and beyond quickly recover the network after a disaster and handle the instant communications by the flash crowd. The drone-cruiser base station can overcome the communications problem for three types of flash crowds, such as in stadiums, parades, and large plaza so that an appropriate number of aerial base stations can be accurately deployed to meet large and dynamic traffic demands. Artificial intelligence can solve these problems by analyzing the collected data, and then adjust the system parameters in the framework of Self-Organizing Network (SON) to achieve the goals of self-configuration, self-optimization, and self-healing. With the help of AI technologies, 5G networks can become more intelligent. This paper aims to provide a new type of service, On-Demand Aerial Base Station as a Service. This work needs to overcome the following five technical challenges: innovative design of drone-cruisers for the long-time hovering, crowd estimation and prediction, rapid 3D wireless channel learning and modeling, 3D placement of aerial base stations and the integration of WiFi front-haul and millimeter wave/WiGig back-haul networks., Comment: 6 pages, 11 figures, accepted by 2019 IEEE Globecom Workshops (GC Wkshps): IEEE GLOBECOM 2019 Workshop on Space-Ground Integrated Networks

This paper proposes a traffic-aware load balancing scheme for machine-to-machine (M2M) networks using software-defined networking (SDN). Load balancing techniques are essential for M2M networks to relieve the heavy loading caused by bursty traffic. Leveraging the capability of SDN to monitor and control the network, the proposed load balancing scheme can satisfy different quality of service requirements through traffic identification and rerouting. Experimental results show that the proposed scheme can reduce service response time up to 50% compared to the non-SDN load balancing scheme.

Multitenant cellular network slicing has been gaining huge interest recently. However, it is not well-explored under the heterogeneous cloud radio access network (H-CRAN) architecture. This paper proposes a dynamic network slicing scheme for multitenant H-CRANs, which takes into account tenants’ priority, baseband resources, fronthaul and backhaul capacities, quality of service (QoS), and interference. The framework of the network slicing scheme consists of an upper-level, which manages admission control, user association, and baseband resource allocation; and a lower-level, which performs radio resource allocation among users. Simulation results show that the proposed scheme can achieve a higher network throughput, fairness, and QoS performance compared with several baseline schemes.

In this paper, we present a hybrid real-time incremental stochastic gradient descent (RI-SGD) updating technique for implicit feedback matrix factorization (MF) recommendation systems. Compared with explicit feedback evaluation scores, implicit feedback data are easier to obtain but pose challenges to MF recommendation systems because of the transformation procedures from raw data to user preference scores. Another challenge for MF recommendation systems is the accuracy issue when the speed of the new input data increases. The proposed RI-SGD is designed for computationally-efficient and accurate time-variant implicit feedback MF recommendation system, which consists of alternating least squares with weight regularization in the training phase and stochastic gradient descent in the updating phase. To demonstrate the advantages of the RI-SGD updating technique in terms of computational efficiency and accuracy, we implement the proposed updating techniques in a real-time music recommendation system. Compared with the method of retraining the entire model, our numerical results show that RI-SGD approach can achieve almost the same recommendation accuracy, but requires only about 0.02% of the retraining time.

In this paper, we present a systemic approach to provide deterministic delay guarantee for dynamic service chaining in software defined networking (SDN). The delay performance of service chaining in SDN is affected by signaling message exchange in control plane and packet transmissions in data plane, respectively. First, we develop an analytical method to characterize the delay performance of control plane when handling traffic with different priorities according to network calculus (NC) and queuing theory. Second, taking into account the estimated delay in control plane, we propose a novel service traversal mechanism to calculate the optimal traversal path for the service chain. We demonstrate that NC delay analysis can provide deterministic quality of service (QoS)-guaranteed service chaining for any specified delay requirements, whereas theoretical queueing delay analysis can only provide statistical QoS guarantee. In summary, the proposed NC delay analysis can help to understand the network design for a future delay sensitive Internet in which deterministic latency must be guaranteed.

This study proposes a multiple mesh-ring (MMR) topology construction with a hybrid routing strategy for Bluetooth low energy (BLE) networks. To achieve the load-balancing network, three design phases, including leader selection, role decision, and scatternet formation, are executed to generate even MMR configurations. First, each master discovers its adjacent slaves to determine a coordinator. Second, the coordinator computes the desired number of piconets with even scatternet link connectivity and distributes the scatternet connection information for each master. Finally, each designated master connects with its associated nodes, including slaves, intrabridges, and interbridges to create the link balanced MMR scatternet. To jointly design an energy-efficient routing protocol for the MMR topology, a hybrid routing strategy is deployed to perform the shortest path routing inside the mesh-ring subnet, and self-routing through different mesh-ring layers. Simulation results demonstrate that the balanced MMR topology leads to a significant gain in terms of network transmission performance and network lifetime when compared to the conventional dual-ring tree (DRT) and cluster-based mesh (CBM) approaches for BLE networks.

This study presents a novel nonuniform power formation algorithm (NUPFA) for Bluetooth low-energy (BLE) networks. With two stages, NUPFA constructs the scatternet in a distributed manner. In the first stage, the master and slave nodes discover each other using low-level power to form an initial scatternet. In the second stage, each master alternatively scans or advertises using normal power to interconnect dual links among any two piconets until the whole scatternet is constructed. Using nonuniform power configuration, an energy-efficient scatternet is generated, since packet transmission power can be reduced in each piconet to minimize overall power consumption. With additional scatternet links, more routing paths can be utilized to reduce the path length and enhance network efficiency. simulation results show that compared with the existing cluster-based on-demand routing protocol (CORP) and Bluetooth Mesh schemes, the NUPFA protocol achieves superior energy efficiency for packet transmission. Also, the dual link features can effectively reduce packet latency, thereby improving the overall performance of BLE mesh networks.

This research presents a new multichain routing strategy named the destination-oriented routing algorithm (DORA). The proposed design generates a new multichain routing scheme to transmit packets for energy-balanced WSNs. Based on the multichain routing, the optimal transmission distance between any two nodes is derived by the mathematical analysis model. With the optimal distance, the design criterion is to select the furthest forwarding node within the communication range and the direction to the sink, thus forming the multichain structures by accurate distance, direction, and multiple paths. With the shorter chain routing, unnecessary energy loss can be reduced to prolong the global network lifespan. The simulation results demonstrate that the proposed DORA doubles the network lifespan, compared with that of conventional PEGASIS and improves 60% lifespan on the RPC protocol.

Thermal ALD (T-ALD) and plasma-enhanced ALD (PE-ALD) processes were employed to prepare the Al 2 O 3 thin films on plasticized polyvinyl chloride (PVC) surfaces at a low deposition temperature of 80 °C. The water contact angle test shows that the T-ALD Al 2 O 3 films exhibited a hydrophobic surface, while the PE-ALD Al 2 O 3 films were hydrophilic. The difference in the surface hydrophobicity is attributed to the carbon residue in Al 2 O 3 films, as revealed by the X-ray photoelectron and Fourier transform infrared spectroscopies. The trend of hydrophobicity follows with the evolution of surface roughness of the T-Al 2 O 3 films, indicating that the surface roughness contributed to the hydrophobicity as predicted by the Wenzel's mode. The result indicates that the surface hydrophilicity/hydrophobicity of the Al 2 O 3 thin films can be modified and controlled by the deposition conditions of the ALD process.

This paper has been presented at : IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (IEEE PIMRC 2019). 8-11 September 2019 Istanbul, Turkey In press / En prensa In moving train networks, two-hop architecture is adopted to improve users experience by reducing the interaction between on-board users and base stations on the train route. In addition, edge networking have emerged as a solution for bringing services to the proximity of the users. However, deploying two-hop and edge networks do not guarantee a continuous service delivery for train users. When a large number of users transit from the train to the land, they experience service interruption due to control signalling storm and backhaul latency. In this paper, we propose a holistic edge service management system to provide mobile service continuity. The contribution of this paper is twofold. First, we develop an enhanced handover scheme that reduces control signals by handling user mobility at the edge. Second, we develop a pre-copy migration scheme that eliminates backhaul latency by relocating containerized applications to the user proximity across edge train networks. Our experimental results show that the two proposed solution can reduce the control signals and migration downtime by 50% and 36%, respectively. This work has been partially funded by the H2020 col-laborative Europe/Taiwan research project 5G-CORAL (grant no. 761586). This research is also partially supported by the Ministry of Science and Technology, under the Grant Number MOST 108-2634-F-009-006 - through Pervasive Artificial Intelligence Research (PAIR) Labs, Taiwan.

Extracting the valuable features and information in Big Data has become one of the important research issues in Data Science. In most Internet of Things (IoT) applications, the collected data are uncertain and imprecise due to sensor device variations or transmission errors. In addition, the sensing data may change as time evolves. We refer an uncertain data stream as a dataset that has velocity, veracity, and volume properties simultaneously. This paper employs the parallelism in edge computing environments to facilitate the top-k dominating query process over multiple uncertain IoT data streams. The challenges of this problem include how to quickly update the result for processing uncertainty and reduce the computation cost as well as provide highly accurate results. By referring to the related existing papers for certain data, we provide an effective probabilistic top-k dominating query process on uncertain data streams, which can be parallelized easily. After discussing the properties of the proposed approach, we validate our methods through the complexity analysis and extensive simulated experiments. In comparison with the existing works, the experimental results indicate that our method can improve almost 60% computation time, reduce nearly 20% communication cost between servers, and provide highly accurate results in most scenarios., 14 pages, 9 figures, Accepted by IEEE Internet of Things Journal

In this letter, we study the on-demand UAV-BS placement problem for arbitrarily distributed users. This UAV-BS placement problem is modeled as a knapsack-like problem, which is NP-complete. We propose a density-aware placement algorithm to maximize the number of covered users subject to the constraint of the minimum required data rates per user. Simulations are conducted to evaluate the performance of the proposed algorithm in a real environment with different user densities. Our numerical results indicate that for various user densities our proposed solution can service more users with guaranteed data rates compared to the existing method, while reducing the transmit power by 29%., Accepted by IEEE Wireless Communications Letters

Because of SDN centralization nature, denial of service (DoS) attacks have become a prominent concern. In the OpenFlow (OF) protocol, the transport layer security (TLS) protocol is recommended to secure the control channel. Unfortunately, the tasks involved in the proper configuration of a secured TLS are very challenging. Even worse, TLS is made an optional mode of communication in OF. As a consequence, some OF-enabled switches and controllers do not adopt TLS. In this paper, we develop a lightweight authentication mechanism, called Hidden Authentication (HiAuth), to protect SDN controller against DoS attacks. HiAuth legitimizes SDN forwarding devices by hiding authentication information into the header of control channel packets. Our experimental results prove that HiAuth is lightweight and can not only mitigate DoS attacks, but also provide high undetectability to the attacker.

With the inertial measurement unit (IMU) and motor technology improvement, the unmanned aerial vehicles (UAVs) become more accessible for different applications. Among the fixed-wing and the rotary- wing UAV, the former is advantageous in having no runway and small craft size which are more attractive to users. The rotary-wing UAVs is used to provide real-time monitor and control on interest targets for ground users. However, the effects of propellers are not addressed in existing rotary-wing UAV channel models. In this paper, we propose a simple air-to-ground radio communication channel for rotary-wing UAVs considering the propellers. Analysis and results are presented based on the model. The results show the maximum of Doppler shifts for all simulated frequencies exceed 300Hz, the requirement of the LTE, and there is potential risks adapt the current communication system directly for serious time-variant multipath and Doppler frequency shift channel issues.

This paper applies Machine Learning (ML) to predict the quality of Air-to-Ground (A2G) links performance for Unmanned Aerial Vehicles Base Stations (UAV-BSs) services. UAV-BSs can instantly identify the status of the current 3D wireless channel in an unknown environment without relying on previous statistical channel modeling. The proposed method that employs the unsupervised learning clustering technology applying to A2G channel modeling in 3D wireless communication scenarios. As environment changing, the proposed method can derive the 3D temporary channel model based on collected RSS data and analyzing. To evaluate the proposed method, the simulation data and measurement data are used to co-verify the performance. As the results shown, the RMSE of conventional statistical channel model and proposed temporary channel model are very similar. The similarity achieves about 91.8% both of the simulation and experimental environments to verify the accuracy and feasibility of our proposed method, and that provides more fast and effective of 3D channel modeling approach.

This paper study the payload data communication of multi-UAVs system, and propose the system architecture. Considering the asynchronous interference of multi-UAVs communication, the filtered-OFDM is introduced to suppress the out-of-band emission and mitigate the inter-subband interference. According to simulation results, the proposed multi-UAVs system based on F-OFDM that is superior to OFDMA scheme in asynchronous interference suppression and spectral efficiency. Finally, we valid the multi-UAVs communication system work well under the considered air-to-ground channel.

Edge networking has become an important technology for providing low-latency services to end users. However, deploying an edge network does not guarantee continuous service for mobile users. Mobility can cause frequent interruptions and network delays as users leave the initial serving edge. In this paper, we propose a solution to provide transparent service continuity for mobile users in large-scale WiFi networks. The contribution of this work has three parts. First, we propose ARNAB architecture to achieve mobile service continuity. The term ARNAB means rabbit in Arabic, which represents an Architecture for Transparent Service Continuity via Double-tier Migration. The first tier migrates user connectivity, while the second tier migrates user containerized applications. ARNAB provides mobile services just like rabbits hop through the WiFi infrastructure. Second, we identify the root-causes for prolonged container migration downtime. Finally, we enhance the container migration scheme by improving system response time. Our experimental results show that the downtime of ARNAB container migration solution is 50% shorter than that of the state-of-the-art migration. This work has been partially funded by the H2020 Europe/Taiwan joint action 5G-DIVE (Grant #859881) and also partially funded by the Ministry of Science and Technology, under the Grant Number MOST 108-2634-F-009-006 - through Pervasive Artificial Intelligence Research (PAIR) Labs, Taiwan.

Paper presented at: IEEE GLOBECOM 2018 Workshops: Intelligent Network orchestration and interaction in 5G and beyond. Abu Dabhi. 9-13 December 2018 In this paper, we present an architecture for transparent service continuity for cloud-enabled WiFi networks called ARNAB: ARchitecture for traNsparent service continuity viA douBle-tier migration. The term arnab means rabbit in Arabic. It is dubbed for the proposed service architecture because a mobileuser service with ARNAB behaves like a rabbit hopping through the WiFi infrastructure. To deliver continuous services, deploying edge clouds is not sufficient. Users may travel far from the initial serving edge and also perform multiple WiFi handoffs during mobility. To solve this, ARNAB employs a double-tier migration scheme. One migration tier is for user connectivity, and the other one is for edge applications. Our experimental results show that ARNAB can not only enable continuous service delivery but also outperform the existing work in the area of container live migration across edge clouds. This work has been partially supported by the H2020 collaborative Europe/Taiwan research project 5G-CORAL (grant num. 761586).

Hydrogen storage on the Ca-decorated defective boron nitride nanosheets (BNNSs) is investigated by first-principle calculation based on density functional theory. Four types of experimentally available defects, B vacancy (VB), N vacancy (VN), Stone-Wales defect (SW) and B–N divacancy (VNB), are considered. It is found that the binding energy of Ca atom on BNNSs with VN and SW defects is smaller than the cohesive energy of bulk Ca. While the Ca atom has strong bonding strength with BNNSs with VB and VBN defects without the problem of aggregation. Six H2 molecules can be absorbed by a Ca atom with the average adsorption energy of 0.206–0.242 eV/H2. The Ca decorated on both sides of BNNSs with VB and VBN defects leads to the hydrogen gravimetric density of 7.6 wt% and 8.0 wt%, respectively. Both the polarization mechanism and the weaker hybridization of Ca-3d orbitals with H-s orbitals are responsible for the adsorption of H2 molecules. The effect of temperature and pressure on the hydrogen storage performance is also investigated, and the results indicate that the hydrogen adsorbed structures of Ca-decorated BNNSs with VB and VBN defects are stable at room temperature under mild pressure.

Hydrogen adsorption and storage performance of Ca-decorated boron nitride nanotubes (BNNTs) with vacancy defects are investigated theoretically using the first-principle calculation. Three types of experimentally available defects, B monovacancy (VB), N monovacancy (VN) and B–N divacancy (VNB), are considered. The Ca atom prefers to reside on the vacancy site of BNNTs-VB and BNNTs-VBN without the problem of aggregation. The H2 adsorption results indicate that six H2 can be adsorbed by a Ca atom with the average adsorption energy slightly larger than 0.20 eV/H2. The hydrogen gravimetric density is 6.4 wt.% and 6.9 wt.% for 8Ca/BNNT-VB and 8Ca/BNNT-VBN systems, respectively. The polarization interaction and the weaker orbitals hybridization between Ca and H2 are responsible for the hydrogen adsorption. The stability of the H2 adsorbed complexes is also investigated by considering the temperature and pressure. The results indicate that the H2 adsorbed structures of Ca-decorated BNNSs with VB and VBN defects are stable at room temperature under commonly used mild pressure.

Network coding is an important cloud storage technique, which can recover data with small repair bandwidth and high reliability compared to the existing erasure coding and replication methods. However, regardless of which data recovery technique is used, the repaired data in a geographically distributed cloud storage system are easy to be eavesdropped at the transmission link between the local datacenter and its remote backup site. This kind of network security issue is called link eavesdropping in this paper. For a network coded cloud storage system, we propose a systematic design methodology to determine the important data recovery system parameters for any specified security level. Through analysis, we present the performance curves to relate the remote repair bandwidth and the number of coded data fragments. Consequently, all the important system parameters of a network coded data recovery system, including the number of storage nodes and the link capacity between the datacenter and the backup site, can be precisely designed for satisfying different security level requirements.

This paper thoroughly explored the fundamental interactions between cell association, cell load and throughput in a green (energy-efficient) small cell network in which all base stations form a homogeneous Poisson point process (PPP) of intensity $\lambda_B$ and all users form another independent PPP of intensity $\lambda_U$. Cell voidness, usually disregarded due to rarity in cellular network modeling, is first theoretically analyzed under generalized (channel-aware) cell association (GCA). We showed that the void cell probability cannot be neglected any more since it is bounded above by $\exp(-\lambda_U/\lambda_B)$ that is typically not small in a small cell network. The accurate expression of the void cell probability for GCA was characterized and it was used to derive the average cell and user throughputs. We learned that cell association and cell load $\lambda_U/\lambda_B$ significantly affect these two throughputs. According to the average cell and user throughputs, the green cell and user throughputs are defined respectively to reflect whether the energy of a base station is efficiently used to transmit information or not. In order to achieve satisfactory throughput with certain level of greenness, cell load should be properly determined. We presented the theoretical solutions of the optimal cell loads that maximize the green cell and user throughputs, respectively, and verified their correctness by simulation., Comment: 15 pages, 8 figures

Pilot contamination (PC) is known as a dominant factor for the performance of time-division duplex massive MIMO systems. Most of the existing studies of PC assumed perfect inter-cell synchronization, which is too costly to achieve or even impossible in practice. In this paper, we consider the unsynchronized scenario, in which the reused pilot signals from other cells are subject to timing misalignment, and propose a method for pilot sequence design for reducing the mean square error (MSE) of the linear least squares (LS) channel estimation. An analytic MSE formula is first derived, which is a very complicated function of the pilot coefficients. To ease analysis, an upper bound on the MSE is then derived. Through minimization of this upper bound, an analytic solution is obtained. Computer simulations are used to illustrate the performance of the proposed pilot sequence.

In this paper, the design of a smart headband for epileptic seizure detection is presented. The proposed headband consists of four key components: 1) an analog front-end circuitry; 2) an epileptic seizure detection tag (ESDT); 3) a Bluetooth low-power chip; and 4) customized electrodes. All the above components are integrated into a fabric headband with only 50.3 g. The smart headband system dissipates 55.89 mW. The epileptic seizure detection algorithm inside ESDT is validated by using Boston Children’s Hospital’s CHB-MIT scalp EEG clinical database with the detection rate of 92.68% and the false alarm of 0.527/h. We develop a service APP connected to the cloud so that the patients’ health condition can be recorded and then referenced by doctors for further diagnosis or research.