Arash Maskooki

Opportunistic routing for body area network

Wireless body area network (WBAN) is an interesting application of sensor networks which can revolutionize our interaction with the outside world. WBAN like any other sensor network suffers limited energy resources and hence preserving the energy of the nodes is of great importance. Unlike typical sensor networks WBANs have few and dissimilar sensors. In addition, the body medium has its own propagation characteristic which means that the existing solution for preserving energy in wireless sensor networks might not be efficient in WBANs. The quality of the links between the nodes in WBAN is changing frequently due to the moving nature of the body. This can pose a major problem especially in the energy efficiency merit. In this work we have proposed an opportunistic scheme to exploit the body movements during the walking to increase the life time of the network. The results show that comparing to the existing methods this work can increase the life time of the network.

Frequency Domain Skin Artifact Removal Method for Ultra-Wideband Breast Cancer Detection

Using ultra-wide band (UWB) microwave pulse for breast cancer detection has been greatly investigated recently since it does not expose the patient to any harmful radiation and the implementation is relatively cheaper than other methods such as MRI or X-ray. An issue in UWB imaging of breast cancer is the strong backscatter from the breast skin which is in orders of magnitude larger than the pulse backscattered from the tumor and should be eliminated before processing the signal for the tumor detection and imaging. At present no existing method can efiectively remove this artifact without introducing corruption to the tumor signature. In this paper, a novel method to eliminate this artifact is proposed which employs a frequency domain model to isolate and remove skin related information from the signal. This method is compared with the existing methods of the skin artifact removal in difierent scenarios. The results show that the new method can overcome the shortcomings of the previous methods and improve the detection of the tumor in the sense of the tumor to clutter response ratio.

Adaptive Routing for Dynamic On-Body Wireless Sensor Networks

Energy is scarce in mobile computing devices including wearable and implantable devices in a wireless body area network. In this paper, an adaptive routing protocol is developed and analyzed which minimizes the energy cost per bit of information by using the channel information to choose the best strategy to route data. In this approach, the source node will switch between direct and relayed communication based on the quality of the link and will use the relay only if the channel quality is below a certain threshold. The mathematical model is then validated through simulations which shows that the adaptive routing strategy can improve energy efficiency significantly compared with existing methods.

A Novel Approach to Joint Flexion/Extension Angles Measurement Based on Wearable UWB Radios

In this paper, a new method for measuring and monitoring human body joint angles, which uses wearable ultrawideband (UWB) transceivers mounted on body segments, is proposed and investigated. The model is based on providing a high ranging accuracy (intersensor distance) between a pair of transceivers placed on the adjacent segments of the joint center of rotation. The measured distance is then used to compute the joint angles based on the law of cosines. The performance of the method was compared with a flexible goniometer by simultaneously measuring joint flexion-extension angles at different angular velocities, ranging between 8 and 90°/s. The measurement errors were evaluated by the average differences between two sets of data (ranging from 0.8° for slow movement to 2.8° for fast movement), by standard deviation (ranging from 1.2° to 4.2° for various movement speeds) and by the Pearson correlation coefficient (greater than 0.99) which demonstrates the very good performance of the UWB-based approach. The experimental results have shown that the system has sufficient accuracy for clinical applications, such as rehabilitation.

Using wearable UWB radios to measure foot clearance during walking

Foot clearance above ground is a key factor for a better understanding of the complicated relationship between falls and gait. This paper proposes a wearable system using UWB transceivers to monitor the vertical heel/toe clearance during walking. First, a pair of very small and light antennas is placed on a point approximating to the heel/toe of the foot, acting as a transmitter and receiver. Then, the reflected signal from ground is captured and propagation delay is detected using noise suppressed Modified-Phase-Only-Correlator (MPOC). The performance of the UWB-based system was compared with an ultrasound system for stationary movements. The experimental results show that an overall mean difference between these two systems is about 0.634mm with correlation coefficient value of 0.9604. The UWB-based system is then used to measure foot clearance during walking which shows promising results for gait events detection.

An Accurate 3D UWB Hyperbolic Localization in Indoor Multipath Environment Using Iterative Taylor-Series Estimation

An accurate approach in locating a source based on the intersections of hyperboloids defined by the time difference of arrival (TDOA) of a signal received at a number of reference nodes is proposed and investigated. The demonstrated system consists of a UWB transmitter, irradiating a UWB Gaussian pulse, and eight coherent receivers where positions are known. Techniques like optimal template-based estimator used for time delay estimation are discussed. The effect of errors in range measurement on the accuracy of positioning is reduced by adopting Taylor-series-based minimization procedures. Furthermore, we evaluate the performance of the proposed method by comparing with traditional LSE minimization procedure in both simulation and experimental results. The experimental results show that the target positioning error is only about (0.7198±0.271)cm (mean±standard deviation), which is accurate for potential applications such as robot navigation, personal tracking and monitoring.

Ultra-Wideband Real-Time Dynamic Channel Characterization and System-Level Modeling for Radio Links in Body Area Networks

Reliable on-body wireless network design needs to incorporate the natural body movements. However, existing models are static or pseudo-dynamic models. In this work, the lack of a comprehensive channel model for a moving body is addressed. A dynamic radio channel around the human body is statistically analyzed and modeled. Important parameters of the channel are estimated from measured data and compared with the model prediction. Simulations show that the model can predict the parameters of a system working in a dynamic body area network with 92% accuracy.

Measurement of knee flexion/extension angle using wearable UWB radios

This paper proposes a wearable system using UWB transceivers to measure the knee flexion/extension angle parameter, who is known to be of clinical importance. First, a pair of very small and light antennas is placed on the adjacent segments of knee joint. Then, the range data between these two antennas is acquired using Time of Arrival (TOA) estimator. We further use the measured distance to compute the flexion/extension angle using the law of cosines. The performance of the method was compared with a flexible goniometer by simultaneously measuring knee flexion-extension angle. The experimental results show that the system has reasonable performance and has sufficient accuracy for clinical applications.

Frequency Domain Breast Lesion Classification Using Ultra-Wideband Lesion Response

Shape of an object has been shown to affect the backscattered ultra-wideband (UWB) pulse in frequency domain. Frequency-domain response of the backscattered signal is the sum of damping harmonics where each harmonic is associated with a specific scatterer and the damping factor is a function of the shape of the corresponding scatterer. Backscattered signals from benign and malignant breast tumors are obtained through numerical analysis. The damping factors were extracted from frequency response of the backscattered signal and neural networks were used to classify breast lesions into benign and malignant tumors. Different architectures of neural networks were investigated and cascaded Kolmogorov network was found to be most accurate classifier. Results show that the cascaded Kolmogorov network classifier can increase the accuracy of diagnosis up to 75.2 %, which is higher than existing methods.