A common problem of hearing aids is whistling caused by feedback from the loudspeaker back to the microphone. A method of reducing the negative effects, caused by the feedback, is called feedback cancellation. A variant of feedback cancellation uses a probe signal, which is applied to the speaker of the hearing aid and is used to continuously estimate the feedback. Oticon A/S has suggested a master's thesis with the purpose of designing and evaluating an algorithm generating a probe signal for feedback cancellation systems. The challenge was to find an inaudible probe signal with as much energy as possible. Two approaches have been investigated for generating a probe signal. In the first approach the psychoacoustic principle of masking was used to estimate how much noise that could be added to a signal without being heard. Psychoacoustic models, including masking, are used in MPEG (Moving Pictures Expert Group) audio coding and one of these models has been examined in the thesis. In the second approach a standard LPC (Linear Prediction Coding) algorithm was used. In both the MPEG and the LPC approach, warped signal processing has been utilized improving the methods. A listening test was performed, evaluating the methods generating the probe signal. The purpose of the test was to determine whether the noise, generated using the MPEG and LPC approach, was inaudible. A hearing aid system with feedback cancellation, using the probe signal, was also simulated. The listening test showed that the noise (probe signal) had to be lowered, much more than expected, to be inaudible. As a consequence, shown in the simulations, the feedback cancellation system, using the probe signal, had trouble identifying the feedback of the hearing aid.

Because of the need to move large amounts of data effienciently, optical based communications are a critical component of modern telecommunications. And as a key enabler of optical communications, electrical components play a critical role in optical data links. Optoelectronic integrated circuits provide the bridge between the optical and electrical realms. Electronic integrated circuits are also integral parts of the optical link, interfacing with post processing circuitry and compensating for any limitations along the link. In this investigation, three circuits for optical data link applications are studied. Two optoelectronic integrated circuit front-ends for freespace and long haul applications, respectively and an active filter for near end cross talk cancellation associated with high data rate transmission. The first circuit is an 8x8 monolithic receiver array for a Spatial Division Multiplexing optical link. A compact and low power 8x8 array was designed and demonstrated a channel that received data at rates of 1Gb/s. It is the first completely monolithic demonstration of a 2D receiver array within a conventional ion implanted GaAs MESFET process. The second circuit demonstrated a long wavelength (1.55 m) optoelectronic receiver for long haul applications. The circuit utilized a TWA topology, which maximizes the available bandwidth from the GaAs MESFET process. It incorporated a thin-film inverted MSM photodetector to achieve nearly monolithic integration. The final circuit is a tunable high pass active filter in 0.18 m CMOS technology. As part of a NEXT noise canceller architecture, it will provide the means to extend data transmission in FR-4 legacy backplanes into the tens of Gb/s datarate.

A combined solution of the Feed-Forward Equalizer (FFE) and Near-End Crosstalk (NEXT) noise cancellation technique was suggested. The techniques increase data throughput and improve link quality in the 20-in FR4 legacy backplane application. Backplane channel loss and coupling noise were measured and characterized to develop the corresponding behavioral channel model. The receiver-side FFE with 4-tap Finite Impulse Response (FIR) filter structure was adopted as the optimum equalizer topology. The 4-tap FIR filter consists of tap delay line with tap-spacing 33 ps and linear tap-gain amplifiers. The tap coefficients were calculated with the Minimum-Mean-Squared-Error (MMSE) algorithm. A 0.18-um CMOS 4-tap FIR filter IC was designed and fabricated. The experiment results showed the 20-Gbit/sec 4-PAM and 10-Gbit/sec NRZ signal were successfully equalized for the 20-in FR4 legacy backplane channel. Moreover, the suggested NEXT noise cancellation technique consists of coarse- and fine-cancellation stages. The 0.18-um CMOS building block ICs such as 7-tap FIR filter, tunable active Pole-Zero (PZ) filter, and a temporal alignment delay line were fabricated. The experiment results showed that 6-dB Signal-to-Noise Ratio (SNR) improvement was achieved by the developed NEXT noise cancellation technique.

In this thesis, we investigate a wireless communications system with distributed transmit antennas. Under such system scenario, the received signal has multiple carrier frequency offsets (CFOs) since each transmitter has its own oscillator, leading to serious inter-carrier interference (ICI) at the receiver end. Therefore, an ICI self-cancellation scheme is proposed in this thesis, where two different relay nodes use different sub-carriers. When the signals from different relay nodes are combined at the destination node, the ICI self-cancellation can be achieved. In addition, the quality of the received signal can be further improved if the residual CFO can be properly compensated. Traditionally, the medium value of the various CFOs is taken for compensation because of its simplicity. However, a medium value does not result in the optimal performance. In this thesis, a close form expression of optimal CFO is derived to maximize the average signal to interference power ratio. It is shown that the optimal CFO compensation is a function of channel state and individual CFOs. Simulation experiments are conducted to investigate the performance of the proposed scheme. It is shown that the system bit error rate can be substantially improved when the CFO is less than 0.3 subcarrier spacing.

This thesis presents a feedback design for an active noise cancellation (ANC) headphone. The designed ANC headphone consists of an analog controller, an audio power amplifier, a headphone speaker, a mini microphone, and a microphone amplifier, which constitute a feedback loop. The controller design follows the method of feedback sensitivity shaping with degree constraint introduced by R. Nagamune and A. Blomqvist in 2005. The advantage of this method is that it eliminates the needs for choosing an analytic weighting function and performing model reduction to yield a lower-order controller, as commonly required in conventional H2/H¡Û optimizations. A fifth-order analog controller for the ANC headphone is designed. The experimental result shows a maximum acoustic noise reduction of 19.7 dB near 200 Hz and an overall noise reduction of more than 10 dB in the control frequency band from 107 Hz to 523 Hz. Moreover, the out-of-band noise amplification is limited to a barely noticeable level of 4.26 dB.

Ultra¬wide band (UWB) system is a breakthrough in wireless communication, as it provides data rate one order higher than existing ones. This dissertation focuses on the design of CMOS integrated frequency synthesizer and its building blocks used in UWB system. A mixer¬based frequency synthesizer architecture is proposed to satisfy the agile frequency hopping requirement, which is no more than 9.5 ns, three orders faster than conventional phase¬locked loop (PLL)¬based synthesizers. Harmonic cancela¬tion technique is extended and applied to suppress the undesired harmonic mixing components. Simulation shows that sidebands at 2.4 GHz and 5 GHz are below 36 dBc from carrier. The frequency synthesizer contains a novel quadrature VCO based on the capacitive source degeneration structure. The QVCO tackles the jeopardous ambiguity of the oscillation frequency in conventional QVCOs. Measurement shows that the 5¬GHz CSD¬QVCO in 0.18 µm CMOS technology draws 5.2 mA current from a 1.2 V power supply. Its phase noise is ¬120 dBc at 3 MHz offset. Compared with existing phase shift LC QVCOs, the proposed CSD¬QVCO presents better phase noise and power efficiency. Finally, a novel injection locking frequency divider (ILFD) is presented. Im¬plemented with three stages in 0.18 µm CMOS technology, the ILFD draws 3¬mA current from a 1.8¬V power supply. It achieves multiple large division ratios as 6, 12, and 18 with all locking ranges greater than 1.7 GHz and injection frequency up to 11 GHz. Compared with other published ILFDs, the proposed ILFD achieves the largest division ratio with satisfactory locking range.

A non-coherent PN code tracking loop with multi-user detection and simplified multi-path interference cancellation (MPIC) is proposed for direct sequence spread spectrum communications system. A decorrelator decision-feedback detector (DDFD) is first applied on the incoming signal to mitigate the multi-user interference. Then, a simplified multi-path interference cancellation (MPIC) is further used to increase signal quality. Finally, a modified code tracking loop (MCTL) is adopted for non-coherent PN code tracking. Mathematical expressions of the S-curve and tracking jitter are derived. Mean time to lose lock is also compared with traditional tracking loops. From the numerical results, we know the proposed PN code tracking loop can efficiently mitigate the interference from multi-user and multi-path and improve the performance of code tracking loop.

This thesis proposes a modified inductively degenerated common-emitter low-noise amplifier. To add a series-shunt feedback capacitance in series to the base of the cascode transistor for increasing the load impedance of the common-emitter transistor and enhancing the Miller effect, it is applied to improve the circuit¡¦s performance. By thoroughly studying the Miller effect for the input matching, noise, and linearity analysis and derivation of the modified structure, the theoretical analysis and experiments demonstrate the improved linearity and well noise performance. In addition, the proposed method is presented with the good figure of merit. The proposed method is presented in a hybrid circuit with the NEC 2S5010 NPN transistor for 900 MHz applications. It demonstrates that this method improves the linearity and the figure of merit has been increased by 50 to 70 percent. Moreover, the novel low noise amplifier is designed with a 0.35£gm SiGe BiCMOS process supported by the TSMC for 5.7 GHz WLAN band applications. It is found that the circuit has the characteristic of IM3 nonlinearity cancellation because the cascode transistor eliminates the third-order intermodulation genaerated by the common-emitter transistor. This thesis establishes a realizable method for high-linearity low-noise amplifier.

<p>Detta examensarbete utröner möjligheten att aktivt släcka ut en amplitudstark frekvensmodulerad sinussignal digitalt. Detta skall göras utan förkunskap om signalen gällande fas, frekvens eller amplitud. Eventuellt övriga, men betydligt svagare, sinussignaler i den analoga signalen skall inte påverkas nämnvärt av utsläckningen. För att lösa problemet har olika teorier för att lösa delproblemen med att mäta amplitud, fas och frekvens studerats varefter en lämplig kombination valts för implementering i en FPGA. Prestandan som uppmätts har jämförts med vad som teoretiskt är möjligt och vi konstaterar att metoden löser problemet på ett adekvat sätt men är i behov av en del förbättringar som diskuteras i slutet.</p> / <p>This master thesis investigates the possibility to cancel a frequency modulated sinusoid with high amplitude in the digital domain. This will be done without any pre-knowledge of the signal regarding phase, frequency or amplitude. Other arbitrary, but weaker, sinusoids present in the analog signal should not be noticeable affected by the cancellation. To solve the partial problems of measuring amplitude, frequency and phase, different theories have been studied and thereafter a suitable combination was chosen for implementation in a Field Programmable Gate Array, FPGA. The performance measured has been compared to the theoretical maximum performance and it has been concluded that the proposed implementation solves this problem adequately but needs some enhancements, these are discussed in the final chapter.</p>

Thesis (doctoral)--Universität der Bundeswehr, München, 2000. / Includes bibliographical references and index.