<p>In this thesis, we consider a wireless communication system which has several transmitters (users), and one receiver (base station). The goal is to extract the data from the desired transmitter, we have to combat noise and interferences from the other transmitters (users). In order to decrease the probability of error we use space diversity; therefore the desired transmitter (user) and receiver (base station) are equipped with arrays of antennas.</p> <p>We use Orthogonal Space-Time Block Codes (OSTBC) because they reach full diversity-gain. Furthermore, these codes enable us to introduce a new linear-receiver, which suppresses the interference and decodes the received data simultaneously.</p> <p>The proposed receiver is designed by enforcing an underline operator on the vector of the received data. We minimize the output power subject to a constraint which preserves a unity gain for all symbols in the desired transmitter's block of data, at the same time, the intra block interference (self-interference) for the desired transmitter (user) is completely suppressed.</p> <p>Simulation results show that the proposed method has substantially lower symbol error rate compared to the matched filter receiver and linear receiver presented by Naguib and Seshadri. Note that in absence of interference, our receiver simplifies to the matcher filter, which is the Maximum Likelihood receiver for a single user MIMO system using Orthogonal Space-Time Block Codes.</p> / Master of Applied Science (MASc)

<p>In this thesis, we reexamine the problems of computer-aided classification and pairing of human chromosomes. Traditionally researchers have dealt with the problem of classification and pairing separately. In our work, we propose to jointly optimize the solutions of these two very closely related problems. The combined problem is formulated into one of optimal three-dimensional assignment with an objective function of maximum likelihood. This formulation poses two technical challenges: 1. estimation of the posterior probability that two chromosomes form a pair and the pair belongs to a class, and 2. good heuristic algorithms to solve the three-dimensional assignment problem which is NP-hard. In our work, we present various techniques to solve these problems. We also generalize our algorithms to cases where the cell data are incomplete as often encountered in practice.</p> / Master of Applied Science (MASc)

<p>To overcome optical signal attenuation m fiber-optic transmission system, amplifiers must be used in the optical link. However, the accompanying amplified spontaneous emission (ASE) noise in optical amplifiers becomes the major factor that impairs the amplified optical signal, thereby degrade the system performance especially in a cascaded amplifier system. This thesis aims at providing a numerical tool that simulates the noise effects when applying optical amplifiers in high-speed long haul transmission systems. A time domain model for Erbium-doped fiber amplifier (EDFA) and semiconductor optical amplifier (SOA), which is capable of handling the locally generated broadband ASE noise and multichannel signals, is developed with affordable computational complexity. Not only the averaged ASE noise contribution to gain saturation has been included in this model, but also the random nature of the ASE noise in optical amplifiers is incorporated through appropriate statistics. Other effects such as the nonuniform spatial distribution of carriers and frequency chirping are also considered in this model. This model is implemented numerically in order to examine the static and dynamic behaviors of the optical amplifiers. Various effects arising from the ASE noise are shown in signal waveforms picked up at any point along the optical link. Furthermore, the newly developed optical amplifier simulator is integrated with an existing time-domain simulation platform that has many other optical components assembled and is capable of handling point-to-point multichannel fiber-optic transmission system in arbitrary configuration. Consequently, the influence of the ASE noise on power booster, in-line amplifier, preamplifier, and the combination of them has been studied thoroughly in both single-channel and WDM fiber-optic transmission systems.</p> / Master of Applied Science (MASc)

<p>With the rapid increase in the market of portable wireless devices, the life time of the battery has become very important, especially in wireless medical telemetry, where it becomes very critical due to the nature of the application. As part of the wireless system, RF circuits are required to have low-power consumption as one of the ways to increase the device life time before the need to charge or change its battery. Meanwhile, the low-power RF circuit should maintain the basic noise and linearity performance required to guarantee an acceptable BER (Bit-Error Rate) at the receiver end. Furthermore, these circuits must be cheap and compatible with digital technologies.</p> <p>This thesis focuses on the design of CMOS voltage-controlled oscillators (VCO), as one of the building blocks ofa wireless transceiver. Power consumption reduction of the VCO is the main objective in the thesis. The first step to achieve the low-power requirement is to optimize the quality factor of the passive devices, such as inductors and capacitors. Therefore, we have introduced a new technique to enhance the quality factor of the monolithic inductors. This technique is based on decreasing the overall parasitic resistance while increasing the mutual inductance between the strips of the inductor. Momentum simulations are performed to compare between the proposed technique and another regular inductor that has the same inductance value. The proposed inductor shows 23% enhancement in its quality factor.</p> <p>Three ultra low-voltage yeas are presented. The first two yeas are designed to operate at frequencies equal to 600 MHz and 2.4 GHz. The novelty in these designs is the low-supply voltage with a value of 400 mY, which is about 100 mV below the threshold voltage ofthe transistor. Biasing in sub-threshold, leads to reducing the power consumption performance. These oscillators are the first RF oscillators to be biased in sub-threshold.</p> <p>The third oscillator was designed to operate at 6 GHz. In this design we have added another contribution, where we forward biased the body contact of the VCO to overcome the high frequency effects that occur in sub-threshold. The three oscillators are designed to operate with supply voltages of 350 mV, 400 mV and 450 mV, respectively. The oscillators have a minimum phase noise of -128 dBc/Hz, -123 dBc/Hz and -113 dBc/Hz at 1 MHz offset, respectively, and power consumption of 36 μW, 41 μW and 90 μW, respectively. These results show the minimum power consumption and supply voltage operation reported in the literature. In the analysis of these oscillators, discussions are given on the large signal performance of the oscillator and the effect of the inductor mismatch on the power consumption and the phase noise performance. From these three veos, the 600 MHz oscillator was measured. It operates with a supply of 430 mV, consumes 257 μW and has phase noise of -132.5 dBc/Hz at 600 kHz offset.</p> <p>To show the effectiveness of the low-power technique used with the above Yeas, another oscillator is designed at 2.4 GHz with a higher supply voltage of 1 V. The oscillator's phase noise is -123.1 dBc/Hz and it consumes 135 μW which is considered very good compared to other oscillators operating with the same supply voltage. Tuning is realized by using body-biasing of the oscillator. Therefore, we have presented a unique discussion on the effect of body-biasing on the performance ofthe oscillator. Measurments for this oscillator was obtained, however, it does not match the simulations due to the package parastics.</p> / Master of Applied Science (MASc)

<p>As the dimensions of the transistors keep decreasing, the effects of transistors' dimensions fluctuation will become increasingly important for differential radio-frequency (RF) integrated circuits (ICs). In most RFICs, the transistor's channel length is usually much smaller than its channel width. Therefore, small unintentional changes in the channel length due to unavoidable fabrication process variations can cause a large circuit performance deviation, and thus, the effects of channel length fluctuation are extremely important.</p> <p>This thesis investigates the effects of channel length fluctuation on the performance of RF oscillators. Two fully integrated, cross-coupled, differential LC oscillators have been designed in triple-well 0.18 J.lm CMOS technology and used as the test circuits. The measured results of the first VCO has a relatively large frequency tuning range of ~25% because two pairs of varactor are used and a power consumption <em>P</em>_con less than 1.5 m W with 1.5 V supply voltage. Also, it has an output power <em>P_out</em> of -13.3 dBm and a <em>P_con</em> of 1.31 mW at an oscillation frequency <em>f_o</em> of 5.5 GHz, which corresponds to the best phase noise performance of -121.2 dBclHz at 1 MHz offset frequency.</p> <p>The simulated results of the second oscillator has a <em>f_o</em> of 1.12 GHz, a <em>P_out</em> of 1.79 dBm and a <em>P_con</em> of 6.479 mW with 1.8 V supply voltage. Both the <em>f_o</em> and <em>P_out</em> are found to be decreased (increased) when the fabricated channel length is larger (smaller) than the one at the design and simulation stages. The maximum variation of <em>f_o</em> and <em>P_out</em> are 2.45 kHz and 0.08 dBm, respectively, with less than 20% of variation in the channel length. Formulae have been derived to predict these variations. Also, the use of body bias on the transistors is proposed as a means to compensate for the changes in the frequency performance characteristics due to the channel length fluctuations. In addition, it is found that the variation of <em>f_o</em> and <em>P_out</em> due to channel length fluctuations is more significant in oscillators with higher <em>f_o</em>. A recommendation is given to designers of how to design an oscillator with high <em>f_o</em> while minimizing the sensitivity of <em>f_o</em> and <em>P_out</em> to channel length fluctuations. While the phase noise is expected to be affected by the channel length fluctuations, simulation result does not show a significant dependence of the phase noise on channel length. This discrepancy is believed to be caused by the inaccuracy of the phase noise simulation.</p> / Master of Applied Science (MASc)

<p>Thanks to geometry scaling, CMOS is becoming the technology of choice for the implementation of radio-frequency and microwave integrated circuits. While CMOS has several advantages over other technologies, such as low-cost and the possibility to integrate analog and digital circuitry on the same chip, its use for high-frequency analog circuits also presents several challenges, because there are some areas where scaling has impaired instead of improving the active and passive device performance. While several techniques can be used to minimize these undesirable effects, many of them only work over very narrow frequency bands; the implementation of circuits that achieve a desired performance over a very wide frequency band is thus a major challenge. Moreover, with further reduction of the transistor dimensions, new effects, such as gate current due to quantum-mechanical tunneling through the gate oxide, will become increasingly significant.</p> <p>This thesis deals with the analysis and design of CMOS broadband amplifiers. A distributed amplifier in a 0.18 μm standard CMOS technology was designed, implemented and measured. It achieves a bandwidth of 2-13 GHz with a 6 dB gain, and better than -9 dB input and output reflections, while consuming 86 mW from a 1.8 V supply and using 2.6 x 1.3 mm² of chip area. The variation of the amplifier characteristics with temperature was studied in the range from 25°C to 125°C. It was found that the forward gain and noise figure change significantly with temperature, while the reflection coefficients, reverse gain and group delay are largely unaffected.</p> <p>A resistive-match amplifier was implemented in 0.18 μm CMOS technology. It has an average gain of 6.5 dB in the 2-7 GHz band with a noise figure lower than 5.4 dB, with input reflection coefficient of less than -3.5 dB and output reflection coefficient of less than -5 dB. It consumes 18.4 mW from a 1.8 V supply, and occupies 1.28 x 0.55 mm² of chip area. It was found that the gain reduction with temperature is much smaller than that of the distributed amplifier.</p> / Master of Applied Science (MASc)

<p>Muscle atrophy is a common problem among patients with end-stage renal disease (ESRD). This is, in part, due to low exercise capacity and inactivity. Therefore, neuromuscular electric stimulation (NMES) is proposed to alleviate this problem. NMES has been shown to elicit strength gains in other diseased populations. Muscle stimulators on the market today are unable to provide the required control and channels necessary to test the above hypotheses. Thus, a multi-channel, computer controlled muscle stimulator was developed. The system was tested through a series of three pilot studies. First, the safety of NMES was verified. Secondly, the system underwent clinical testing to determine if it is possible to increase muscle strength within the ESRD population. Finally, the idea of incorporating depolarizing pulses to the stimulus pulse train was investigated in an attempt to minimize perceived pain.</p> <p>Results show that NMES is a safe and well tolerated form of rehabilitation. The system was shown to be capable of enduring the rigors of testing. Clinically, NMES showed an increase in peripheral muscle strength within the ESRD population that are comparable to the results shown by previous NMES studies involving other diseased populations. Also, the incorporation of prepulses proved to be an effective method of reducing the perceived pain of NMES.</p> <p>The effectiveness of NMES on ESRD patients should be evaluated further. This includes the addition of more subjects to the study in order to show a significant increase in strength.</p> / Master of Applied Science (MASc)

<p>This thesis develops a marginalized particle filtering algorithm for the blind system identification problem. The blind system identification problem arises in many fields, including speech processing, communications, biomedical signal processing, sonar and seismology. The state space model under consideration uses a time-varying autoregressive (AR) model for the sources, and a time-varying finite impulse response (FIR) model for the channel. The multi-sensor measurements result from the convolution of the sources with the channels in the presence of additive noise. A numerical approximation to the optimal Bayesian solution for the sequential state estimation problem is implemented using the particle filter. Estimates of the sources are recovered directly by marginalizing the AR and FIR coefficients out of the posterior distribution for the unknown system parameters. The resulting marginalized particle filtering algorithm allows efficient identification of the system. Simulation results are given to verify the performance of the proposed method. The block sequential importance sampling (BSIS) formulation of the particle filter is also introduced to exploit the structure inherent in the convolution state space model.</p> / Master of Applied Science (MASc)

<p>IEEE 802.11 wireless local area networks (WLANs) are increasingly used to support real-time services such as voice and video. Reliable and portable operation, however, is often difficult due to factors such as imperfect customer access point (AP) installation, unpredictable WLAN coverage, and unexpected co-channel interference.</p> <p>There has been much recent activity that considers the combination of ad hoc relaying and infrastructure wireless networks as alternative solutions to coverage extension. In this thesis, we propose and further investigate the use of IEEE 802.11 Ad Hoc Assisted Handoff (AAHO). In AAHO, an additional ad hoc hop may be used by a mobile station (MS) to obtain the range extension or channel quality needed to maintain its real-time voice connection. An MS that is currently not carrying active traffic may offer itself as a potential relay station (RS). Three versions of IEEE 802.11 AAHO are discussed in this thesis. In the case of Backward Ad Hoc Assisted Handoff (BAAHO), the additional hop uses a relay station which already has an IEEE 802.11 association with the AP that the MS is using. In the case of Forward Ad Hoc Assisted Handoff (FAAHO), however, the additional hop uses a relay station whose AP is different from the one that the MS is currently using. Hybrid Ad Hoc Assisted Handoff (HAAHO) is a combination of these two concepts that allows an MS to perform in either BAAHO or FAAHO mode. The proposed AAHO schemes are backward compatible to existing IEEE 802.11 infrastructure. They can be implemented as a transparent overlay across existing IEEE 802.11 deployments. Two implementation options for achieving this compatibility are proposed. A criterion in selecting relay station is introduced, which permits mobile stations to control real-time relaying of voice packets. Our simulation results show that AAHO can greatly improve performance in many realistic situations.</p> / Master of Applied Science (MASc)

This thesis studies throughput improvement for TCP traffic in IEEE 802.11-based multihop ad hoc wireless networks. Due to the Incompatibility between TCP and the IEEE 802.11 distributed coordination function (DCF) protocols, the reaction of TCP in case of packet losses can significantly reduce TCP end-to-end throughput. In this thesis, we propose an opportunistic link scheduling (OL8), which is a simple enhancement to the IEEE 802.11 DCF protocol and intends to improve the compatibility between TCP and MAC layer protocols in multihop ad hoc networks. With OL8, a link with a good channel condition is allowed to transmit multiple packets consecutively as a burst, while the burst size depends on both physical channel fading and MAC layer collisions. The protocol also includes a mechanism to prevent starvation of nodes with poor channel conditions. An analytical model is developed for a four-hop chain to study the effect of the burst size and TCP congestion window size on the end-to-end transmission throughput in opportunistic link scheduling. Our results show that OL8 can significantly improve the end-to-end transmission throughput, while keeping reasonably low transmission delay. The protocol is easy to implement, and requires slight modifications to the IEEE 802.11 protocol. / Master of Applied Science (MASc)