Department Publications

A theoretical framework is presented for the evaluation of sum ergodic rate of a full-duplex underlay device-to-device network, when it shares the uplink resources of a conventional cellular user. The sum-rate of the full-duplex network is compared with a half-duplex network with equivalent radio frequency hardware complexity. Closed-form approximations are derived for the sum ergodic rate of the systems. Furthermore, the sum-rate performances are investigated for the case when a transmit power constraint is imposed on the underlay network to minimize the interference on the cellular network. The analytical results presented can be used as a tool to identify when full-duplex transmissions are viable in underlay device-to-device networks.

A new trajectory planning method for generating bounded and continuous jerk trajectories in joint space has been developed and tested. In manipulator trajectory planning, reduced jerk trajectories are desired for path tracking and vibration suppression. The proposed interpolation algorithm in this research generates a spline, composed with 5th-order, 3rdorder and 5th-order polynomial segments (5-3-5 spline) which can be used for point-to-point trajectories and trajectories with via points. The generated trajectories are continuously differentiable in position, velocity, acceleration, and has a start and end zero bounded continuous jerk profile. The algorithm allows the user to independently define the position, velocity, acceleration and jerk values at both start and end points, via point positions and velocities. These user definable parameters in the proposed 5-3-5 spline algorithm gives the flexibility for generating …

Automatic stroke recognition of badminton video footages plays an important role in the process of analyzing players and building up statistics. Yet recognizing activities from broadcast videos is a challenging task due to person dependant body postures and blurring of the fast moving body parts. We propose a robust and an accurate approach for badminton stroke recognition using dense trajectories and trajectory aligned HOG features which are calculated inside local bounding boxes around players. A four-class SVM classifier is then used to classify badminton strokes to be either smash, forehand, backhand or other. This approach is robust to noisy backgrounds and provides accurate results for low resolution broadcast videos. Our experiments also reveal that this approach needs relatively fewer training samples for accurate recognition of strokes compared to existing approaches.

This paper studies downlink coverage maximization for cellular networks in which base station (BS) locations are modeled using a spatial Poisson point process, considering three different coverage models, and under constraints on transmit power, BS density and transmit power density. Firstly, the coverage optimization problem is solved analytically for the first coverage model that focuses on noise-limited communication by ignoring interference and random fading effects. This model provides useful insights into the significance of bounded path loss models to obtain meaningful solutions for this problem. The other two coverage models are based on the users' received signal-to-interference-plus-noise-ratio (\sinr) from their associated BSs. For these models, it is shown that the coverage optimization problem can be reduced to a constrained single dimensional optimization problem without any loss of optimality …

Brain-computer interfaces (BCIs) require real-time feature extraction for translating input EEG signals recorded from a subject into an output command or decision. Owing to the inherent difficulties in EEG signal processing and neural decoding, many of the feature extraction algorithms are complex and computationally demanding. Presently, software does exist to perform real-time feature extraction and classification of EEG signals. However, the requirement of a personal computer is a major obstacle in bringing these technologies to the home and mobile user affording ease of use. We present the FPGA design and novel architecture of a generalized platform that provides a set of predefined features and preprocessing steps that can be configured by a user for BCI applications. The preprocessing steps include power line noise cancellation and baseline removal while the feature set includes a combination of linear …

The DV-Hop algorithm is an attractive option for the localization of nodes in a wireless sensor network due to its simplicity. However it suffers from poor accuracy. This paper presents a novel algorithm to enhance the accuracy of this algorithm. The idea is to use the DV-Hop algorithm to estimate positions of the anchors which are already known. The estimation errors so obtained are then minimized through an optimization algorithm which finds the optimum correction to the hop size. The locations of the unknown nodes are then estimated with the modified hop size using least squares estimation. We examine the performance of the algorithm for varying node numbers, anchor ratios and radio range. The results show that the proposed algorithm has better performances in localization precision and stability than the original DV-Hop algorithm.

This paper studies the outage capacity of a network consisting of a multitude of heterogeneous mobile users and operating according to the classical opportunistic beamforming framework. The base station is located at the center of the cell, which is modeled as a disk of finite radius. The random user locations are modeled using a homogeneous spatial Poisson point process. The received signals are impaired by both fading and location dependent path loss. For this system, we first derive an expression for the beam outage probability. This expression holds for all path loss models that satisfy some mild conditions. Then, we focus on two specific path loss models ( i.e. , an unbounded model and a more realistic bounded one) to illustrate the applications of our results. In the large system limit, where the cell radius tends to infinity, the beam outage capacity and its scaling behavior are derived for the selected specific …

A low-complexity 3D FIR filter is proposed for selectively enhancing severely corrupted linear trajectory spatio-temporal signals. The proposed 3D FIR filter is separable and consists of a 3D spatio-temporal wide-angle FIR cone filter between two 2D spatial variable-shift filters. Low complexity of the 3D spatio-temporal wide-angle FIR cone filter is achieved by exploiting the spatial-symmetry of the passband and by employing maximal decimation in the temporal dimension. Compared to existing techniques, the proposed 3D FIR filter provides significant reduction of the computational complexity for similar SINR improvement.

A multibeam filter bank (MBFB) is proposed for recovering the broadband far-field signals received from multiple directions while rejecting interfering signals and noise. The MBFB consists of two sections. The first section employs variable fractional delays and independently steers the directions of the multiple beams. It achieves low complexity by using the Farrow structure. The second section achieves low complexity by maximal decimation methods and by exploiting the spatial symmetry of a 3-D double-frustum-shaped passband. Superior performance and lower complexity compared with previous methods are numerically confirmed for a typical narrow-angle broadband multibeam beamformer of the type that could be used in radio astronomy.

This paper studies the coverage maximization for wireless networks in which base station (BS) locations are drawn from a homogenous spatial Poisson point process, and user locations are arbitrary. A user is covered for communication if its received signal-to-noise-ratio (SNR) is above a given threshold value, and the objective is to maximize the coverage probability under per unit area power density constraints. The resulting optimization problem is solved analytically by making use of the underlying concavity in the objective function when transmissions are impaired only by a power-law bounded path loss. Our results show that the optimal transmit power per BS is independent of the power density constraint, and the solution to the optimization problem represents the Pareto optimal boundary between the power density constraint and the coverage probability. Then, these results are extended to a system in which …

The downlink performance of a two-cell relay enhanced wireless network is investigated. A single relay is shared by the two base stations in the system. The relay is assumed to be placed in a geographic region covered by both base stations. The relay station is operating in the decode-and-forward mode. Closed-form approximations which estimate the exact results extremely closely are derived for the outage probability and the ergodic rate of the system. The impact of finite rate channel state information feedback on the user ergodic rate is examined.

Delay variation or jitter is an inherent feature of packet switched communication networks due to bottlenecks. In communication systems buffering is commonly used to mitigate in real time multimedia applications at the expense of an additional delay. This method is therefore not suitable for control systems implemented over packet switched communication networks due to the additional delay which can make the system unstable. Hence, jitter modelling and management is necessary. This paper focuses on the development and experimental verification of a suitable jitter model for wireless and wired network bottlenecks. The effects of jitter are then investigated by quantifying the performance of a simulated Networked Control System (NCS). According to the results, the reduction in attenuation due to jitter compared to a jitter free benchmark ranges from approximately 0.01-0.04dB depending on the encoding type.

Vehicle tracking and surveillance is an area which is having a considerable attention in the context of security and safety. The detection and tacking of moving vehicles through multiple cameras is considered as a method of vehicle surveillance. This work addresses a problem of detecting and matching vehicles through multiple cameras. The power of GPUs are used to increase the number of video streams which can be processed using a single computer. In the detection process the Gabor filter is used as a directional filter and the SURF is used by the matcher to uniquely represent the vehicle.

A novel wireless sensor network architecture designed for gathering and managing onsite dynamic information in emergency situations is presented. The system can operate in two different states-normal and emergency, where switching between states occurs according to perception from the environment. The main functionalities of the system include perceiving, processing and sharing the gathered information within a wireless sensor network which is deployed in a multi-story indoor environment. Algorithms for sensor localization, network self-configuration as well as data filtering and prediction are studied in the characterized emergency environment. The proposed architecture will act to reduce risk in rescue operations following an emergency as well as generating early warnings of possible emergencies.

Flying and handling qualities are substantially dependent on, and this paper is described it, in terms of the stability and control characteristics of UAV. It is essential to be able to describe and quantify the stability and control parameters completely. It is absolutely essential to understand the relationship between the aerodynamics of the airframe and its stability characteristics to prolong the flight endurance and effective deployment. And this paper is described the stability analysis based on the dynamic model of the twin boom double vertical stabilizer UAV.

The Australian construction industry is identified as one among five industries to receive priority attention under the Australian Work Health and Safety Strategy 2012-2022. Among the recorded construction fatalities in 2012, vehicle incidents accounted for 15% and being hit by moving objects accounted for 10%. Tracking of individuals on-site could provide a technological solution to improve safety. This paper proposes a novel, non-tagging-based, wireless networking solution (called device-free localization) to enhance safety on site. Device free localization techniques use radio signal propagation effects such as multi-path fading and shadowing caused by the obstruction(s). The proposed solution defines zones in the construction site. Zone identification is based on the risk associated with the location map of the site. Defining risk profiles considers the activities associated with the geographic location, such as …

Early detection of hypertension generally requires continuous monitoring of blood pressure levels which is not facilitated by traditional methods such as the cuff, which cannot be used in the normal environment for continuous monitoring due to the regular pressurization of certain body parts. Thus there is a need for non-invasive continuous pressure monitoring mechanism. In this paper we present a relationship between Pulse Transit Time (PTT) and Systolic Blood Pressure (SBP) and Diastolic Blood Pressure (DBP) for continuous pressure monitoring and its experimental validation.

Manual analysis of large volumes of video surveillance footage stemming from the widespread deployment of security cameras is error prone, expensive and time consuming. Despite the commercial availability of software for automated analysis, many products lack third party extensibility, the capability to perform simultaneous event detection and have no provision for anomaly detection in highly dense crowded scenes. We present a plugin based software system for video surveillance applications addressing these shortcomings and achieve realtime performance in typical crowded scenes. Core parameters are computed once per frame and shared among plugins to improve performance by eliminating redundant calculations. A novel multiple pedestrian tracking algorithm is incorporated into the framework to achieve the expected performance. We also propose an improvement to anomaly detection in densely …

Abnormal activity detection plays an important role in many areas such as surveillance, military installations, and sports. Existing abnormal activity detectors mostly rely on motion data obtained over a number of frames to characterize abnormality. However, only motion may not be able to capture all forms of abnormality, in particular, poses that do not amount to motion "outliers". In this paper, we propose two different spatio-temporal descriptors, a silhouette and optic flow based method and a dense trajectory based method which additionally include trajectory shape descriptor, to detect abnormalities. These two descriptors enable us to classify abnormal versus non-abnormal activities using SVM. Comparison with existing methods, using five standard datasets, shows that dense trajectory based method outperforms state-of-the-art results in crowd dataset and silhouette and optic flow based method outperforms …

Device free detection and localization is a promising and a very practical research area where installation of expensive sensors and/or tagging the intruder is not feasible. This paper reports a mesh network developed for device free trespasser detection. We use the Received Signal Strength Indicator (RSSI) measurements of the wireless links within the mesh as the sole sensing mechanism. A fingerprint of RSSI measurements characterizes the environment under normal conditions including multi-path effects. A technique to differentiate RSSI variations due to trespassers from those due to multi-path effects is demonstrated. The results of the experiments conducted in an outdoor environment have verified that the proposed method can be used to detect trespassers successfully. The reliability of the detection is improved by cross-correlating RSSI variations among the multiple links in the mesh. The ability to extend …

Multi-view video has gained widespread popularity in the recent years. 3DTV, surveillance, immersive teleconferencing and free view-point television are few notable applications of multi-view video. Excessive storage and transmission bandwidth requirements are the major challenges faced by the industry in facilitating multi-view video applications. This paper presents efficient tools for coding of multi-view video based on the state of the art single view video coding standard H.265/HEVC (High Efficiency Video Coding). Our approach employs the LDI (Layered Depth Image) representation technique which is capable of compactly representing 3D scene content. We propose techniques and algorithms for LDI construction, view synthesis, efficient coding of LDI layers and associated auxiliary information. Subjective assessments indicate that our approach offers more than 50% reduction in bitrate compared to HEVC …

In this paper, we consider a multi-cell multi-user MISO broadcast channel. The system operates according to the opportunistic beamforming framework in a multi-cell environment with variable number of transmit beams (may alternatively be referred as the transmission rank) at each base station. The maximum number of co-scheduled users in a cell is equal to its transmission rank, thus increasing it will have the effect of increasing the multiplexing gain. However, this will simultaneously increase the amount of interference in the network, which will decrease the rate of communication. This paper focuses on optimally setting the transmission rank at each base station such that a set of Quality of Service (QoS) constraints, that will ensure a guaranteed minimum rate per beam at each base station, is not violated. Expressions representing the achievable region of transmission ranks are obtained considering different …

This paper presents a performance optimized FPGA based design for low rank matrix estimation using robust principal component analysis. The system is constructed using mesh connected processor array for singular value decomposition and specialized architectures for merge sort and matrix multiplication for re-composition of the matrix which is computed iteratively. There are no prior implementations of Robust PCA on FPGA and this work presents at least a 50 times performance increment upto 2048×2048 matrices, compared to a Matlab implementation running on Core i7-3610QM, 2.3 GHz processor with 8GB RAM.

Difficult conditions on outdoor terrains make outdoor autonomy for rovers, a challenging task. The conventional wheel odometry method uses orientation measurements to assume a momentary plane to apply wheel encoder readings. On uneven terrains, this method often gives poor results for position tracking, and therefore rarely used. To improve the conventional odometry motion model, immediate terrain data can be used. This paper proposes a novel state variable extension (SVE) method to establish a connection between state space variables of a terrain rover by combining terrain point clouds with rover kinematics. The simulation results show that when the 2D state variables (x, y, yaw) are known, the 2D state can be extended to its 3D state (x, y, z, roll, pitch, yaw) with minimal error. The proposed SVE method is employed in a particle filter to determine the 2D state variables, which in turn results in achieving …