Lijun Qian

Time Synchronization of Cognitive Radio Networks

In this paper, a novel synchronization protocol is proposed especially for Cognitive Radio (CR) networks called CR-Sync. In a CR network, time synchronization is indispensable because of the requirements for coordinated and simultaneous quiet periods for spectrum sensing, as well as the common understanding of time frame/slot in many CR MAC designs. The proposed CR-Sync achieves network-wide time synchronization in a fully distributed manner, i.e., each node performs synchronization individually using CR-Sync. Contrary to many existing synchronization protocols that do not exploit CR attributes, the proposed protocol takes advantage of the potential multiple spectrum holes that are discovered by CR and distributes the synchronization of different pairs of nodes to distinct channels and thus reduces the synchronization time significantly. Detailed analysis of synchronization error and convergence time are provided. Results show that the proposed CR-Sync out-performs other protocols such as TPSN in CR networks.

Performance bound of ad hoc Device-to-Device communications using cognitive radio

The aim of this paper is to study the achievable throughput of an ad hoc Device-to-Device (D2D) communications with cognitive radio capabilities coexisting with cellular user equipment (UE) in the same macrocell. Specifically, we consider ad hoc D2D systems with non-orthogonal resources sharing instead of D2D with infrastructure support. The main objective is to find out how much throughput a device in a D2D pair can achieve in the presence of another D2D pair and a cellular user over fading channels. A closed-form expression for statistics, the Moment Generating Function (MGF) and Complementary Cumulative Distribution Function (CCDF) of multiple interferers in Nakagami-m fading channels in cellular system are presented. By using these expressions, we derive the device throughput for multiple D2D systems in a cellular system. Furthermore, the upper bound for the probability of false alarm, which is required to achieve a certain throughput is deduced. The results of this paper illustrate how the transmission probability and sensing performance affect the achievable throughput in cognitive D2D systems. In addition, these results serve as guidance for the deployment of cognitive D2D systems without infrastructure support.

A multiclass classification method based on deep learning for named entity recognition in electronic medical records

Research of named entity recognition (NER) on electrical medical records (EMRs) focuses on verifying whether methods to NER in traditional texts are effective for that in EMRs, and there is no model proposed for enhancing performance of NER via deep learning from the perspective of multiclass classification. In this paper, we annotate a real EMR corpus to accomplish the model training and evaluation. And, then, we present a Convolutional Neural Network (CNN) based multiclass classification method for mining named entities from EMRs. The method consists of two phases. In the phase 1, EMRs are pre-processed for representing samples with word embedding. In the phase 2, the method is built by segmenting training data into many subsets and training a CNN binary classification model on each of subset. Experimental results showed the effectiveness of our method.

Community based sensing: A test bed for environment air quality monitoring using smartphone paired sensors

Awareness of our surrounding environment is very important. Traditional stationary air quality and pollution monitoring systems which installed in dedicated locations are usually bulky and expensive, and they may not be able to measure at locations of our interest. In this study, we propose a novel framework of an environment air quality monitoring system using a community-based approach. Leveraging on the high penetration of smartphones and low cost and small form factor of certain sensors that can be carried on a key chain with a Bluetooth module, critical measurements such as air quality can be measured by each sensor carried by a member of a community, and be sent to that persons smartphones, and eventually uploaded to a web portal using a corresponding app. Then the aggregated and processed data at the server side can be visualized and shared among the community members and publish through social networks. In this project, we have designed the architecture of the proposed community sensing system, and implemented the system using commercial off-the-shelf (COTS) Sensordrone paired with Android© smartphones. Our system measure temperature, humidity, pressure, carbon monoxide, and battery charge level in real-time and it will provide location based services to users. Extra sensors such as ozone and radiation sensors can be added as well.

Multisensor change detection on the basis of big time‐series data and Dempster‐Shafer theory

With the proliferation of the Internet of Things, numerous sensors are deployed to monitor a phenomenon that in many cases can be modeled by an underlying stochastic process. The goal is to detect change in the process with tolerable false alarm rate. In practice, sensors may have different accuracy and sensitivity range, or they decay along time. As a result, the sensed data will contain uncertainties and sometimes they are conflicting. In this study, we propose a novel framework to take advantage of Dempster‐Shafer theorys capability of representation of uncertainty to detect change and effectively deal with complementary hypotheses. Specifically, Kullback‐Leibler divergence is used as the metric to find the distances between the estimated distribution with the before and after change distributions. Mass functions are calculated on the basis of those distance values for each sensor independently, and Dempster‐Shafer combination rule is applied to combine the mass values among all sensors. In the case of high conflict in various sensor readings, Dezert‐Smarandache combination rule is applied, and the belief, plausibility, and pignistic probability are obtained for decision making. Simulation results using both synthetic data and real data demonstrate the effectiveness of the proposed schemes.

Distributed spectrum monitoring and surveillance using a cognitive radio based testbed

This paper described the development and implementation of a distributed spectrum monitoring and surveillance testbed for identifying and locating RF signals using the Universal Software Radio Peripheral (USRP). We use a centralized RF trace collection testbed to establish the baseline, and we focus on distributed RF trace collection. Challenges associated with synchronization is identified and candidate solutions are discussed. The distributed testbed was implemented using NI USRPs (293×/295×) with LabVIEW. The complex nature of implementing the distributed case necessitate the choice of LabVIEW because of the versatile features provided. Potential applications are discussed and sample traces are demonstrated.

Experimental study of hierarchical Software Defined Radio controlled Wireless Sensor Network

In this paper, we examine an hierarchical Software-Defined Radio (SDR) controlled Wireless Sensor Network (WSN) testbed built in our Wireless Communications Lab. In this testbed, an hierarchical cluster-based topology is employed to fulfill the needs of energy efficiency and scalability where a group of XBOW MicaZ Sensors/Motes in communication with a Universal Software Radio Peripheral (USRP2) forms a cluster. The USRP2s act as cluster heads to perform data collection using the least interfered channels due to their capability in channel sensing and waveform selection. The USRP2 also serves to extend the transmission range of the sensors and eliminates the excessive overhead required in ad hoc WSN cluster heads. The cluster heads receive the data transmission from the motes on multiple channels and relay the data to the central control on a separate channel. It is expected that this testbed would help the research community in understanding and gathering insightful knowledge about SDR controlled WSNs in a practical context.

Short-term load forecasting in smart grid: A combined CNN and K-means clustering approach

Although many methods are available to forecast short-term electricity load based on small scale data sets, they may not be able to accommodate large data sets as electricity load data becomes bigger and more complex in recent years. In this paper, a novel machine learning model combining convolutional neural network with K-means clustering is proposed for short-term load forecasting with improved scalability. The large data set is clustered into subsets using K-means algorithm, then the obtained subsets are used to train the convolutional neural network. A real-world power industry data set containing more than 1.4 million of load records is used in this study and the experimental results demonstrate the effectiveness of the proposed method.

Survey of wireless big data

Wireless big data describes a wide range of massive data that is generated, collected and stored in wireless networks by wireless devices and users. While these data share some common properties with traditional big data, they have their own unique characteristics and provide numerous advantages for academic research and practical applications. This article reviews the recent advances and trends in the field of wireless big data. Due to space constraints, this survey is not intended to cover all aspects in this field, but to focus on the data aided transmission, data driven network optimization and novel applications. It is expected that the survey will help the readers to understand this exciting and emerging research field better. Moreover, open issues and promising future directions are also identified.

Implementing a distributed volumetric data analytics toolkit on apache spark

The multidimensional array is a fundamental data structure that has been widely used in scientific computing, as well as in many big data analytics applications. Distributed multi-dimensional array has been well studied in the High Performance Computing (HPC) platforms; however, little research has been done in the widely-used big data analytics platforms. In this paper, we present an implementation of Distributed Multi-dimensional Array Toolkit (DMAT) on top of the Apache Spark big data analytics platform. The toolkit supports several fashions for multidimensional array distributions, repartition, transposition, access, and data parallelism with a variety of parallel execution templates. This paper introduces the software architecture and implementations of DMAT, and also studies the performance characteristics of some typical multi-dimensional array operations with different configurations.