by Wong Chun Kau, Jolly. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1992. / Includes bibliographical references (leaves 84-88). / ABSTRACT --- p.1 / ACKNOWLEDGEMENT --- p.3 / Chapter 1. --- INTRODUCTION --- p.7 / Chapter 1.1 --- Land Mobile Radio (LMR) Communications / Chapter 1.2 --- Paramilitary Communications Security / Chapter 1.3 --- Voice Scrambling Methods / Chapter 1.4 --- Digital Voice Encryption / Chapter 1.5 --- Digital Secure LMR / Chapter 2. --- DESIGN GOALS --- p.20 / Chapter 2.1 --- System Concept and Configuration / Chapter 2.2 --- Operational Requirements / Chapter 2.2.1 --- Operating conditions / Chapter 2.2.2 --- Intelligibility and speech quality / Chapter 2.2.3 --- Field coverage and transmission delay / Chapter 2.2.4 --- Reliability and maintenance / Chapter 2.3 --- Functional Requirements / Chapter 2.3.1 --- Major system features / Chapter 2.3.2 --- Cryptographic features / Chapter 2.3.3 --- Phone patch facility / Chapter 2.3.4 --- Mobile data capability / Chapter 2.4 --- Bandwidth Requirements / Chapter 2.5 --- Bit Error Rate Requirements / Chapter 3. --- VOICE CODERS --- p.38 / Chapter 3.1 --- Digital Speech Coding Methods / Chapter 3.1.1 --- Waveform coding / Chapter 3.1.2 --- Linear predictive coding / Chapter 3.1.3 --- Sub-band coding / Chapter 3.1.4 --- Vocoders / Chapter 3.2 --- Performance Evaluation / Chapter 4. --- CRYPTOGRAPHIC CONCERNS --- p.52 / Chapter 4.1 --- Basic Concepts and Cryptoanalysis / Chapter 4.2 --- Digital Encryption Techniques / Chapter 4.3 --- Crypto Synchronization / Chapter 4.3.1 --- Auto synchronization / Chapter 4.3.2 --- Initial synchronization / Chapter 4.3.3 --- Continuous synchronization / Chapter 4.3.4 --- Hybrid synchronization / Chapter 5. --- DIGITAL MODULATION --- p.63 / Chapter 5.1 --- Narrowband Channel Requirements / Chapter 5.2 --- Narrowband Digital FM / Chapter 5.3 --- Performance Evaluation / Chapter 6. --- SYSTEM IMPLEMENTATION --- p.71 / Chapter 6.1 --- Potential EMC Problems / Chapter 6.2 --- Frequency Planning / Chapter 6.3 --- Key Management / Chapter 6.4 --- Potential Electromagnetic Compatibility (EMC) Problems / Chapter 7. --- CONCLUSION --- p.80 / LIST OF ILLUSTRATIONS --- p.81 / REFERENCES --- p.82 / APPENDICES --- p.89 / Chapter I. --- Path Propagation Loss(L) Vs Distance (d) / Chapter II. --- Speech Quality Assessment Tests performed / by Special Duties Unit (SDU)

Radio-frequency filters and oscillators are widely used in wireless communication and high-speed digital systems, and they are mostly built on passive integrated inductors, which occupy a relative large silicon area. This research attempted to implement filters and oscillators operating at 1-5 GHz using transistors only, to reduce the circuits’ area. The filters and oscillators are designed using active inductors, based on the gyrator principle; they are fabricated in standard digital CMOS technology to be compatible with logic circuits and further lower the cost. To obtain the highest operating frequency, only parasitic capacitors were used. Two new active-inductor circuits are derived from this research, labeled allNMOS and all-NMOS-II. The all-NMOS active inductor was used to design high-Q bandpass filters and oscillators, which were fabricated in TSMC’s 0.18-µm digital CMOS process. The highest center frequency measured was 5.7 GHz at 0.20-µm gate length and the maximum repeatably measured Q was 665. 2.4-GHz circuits were also designed and fabricated in 0.40-µm gate length. The all-NMOS-II circuit has superior linearity and signal fidelity, which are robust against process and temperature variations, due to its novel structure. It was used in signal drivers and will be fabricated in commercial products. Small-signal analysis was conducted for each of the active-inductor, filter and oscillator circuits, and the calculated performance matches those from simulations. The noise performance of the active inductor, active-inductor filter and oscillator was also analyzed and the calculated results agree with simulations. The difference between simulation and measured results is about 10% due to modeling and parasitic extraction error. The all-NMOS active-inductor circuit was granted a US patent. The US patent for all-NMOS-II circuit is pending. This research generated three conference papers and two journal papers.

Switching frequency modulation techniques, an inductor current sensing circuit for fast switching converter, and a dual converter are proposed, and the simulation results and experimental results are drawn. The experimental results for monotonic and pseudo-random modulation techniques show that the switching noise peak was effectively reduced as much as -19 dBc. The inductor current sensing circuit accurately tracks the output current of the switching converter that switches up to 30MHz. This current sensing circuit is used to drive the slow converter in the dual converter. The dual converter consists of a fast converter and a slow converter. The fast converter provides only the high frequency conponents in the output current, and the slow converter provides the majority portion of the output current with a higher efficiency. Therefore, the dual converter can have a fast transient response without sacrificing its efficiency. All chips are fabricated in a standard CMOS 0.18um process.

Fade and inter-fade duration data obtained from the three beacons at 12, 20, and 30 GHz aboard the OLYMPUS satellite were analyzed. The different types of signal impairments and their causes were highlighted and a literature survey conducted. Twelve months of fade and inter-fade data were analyzed and the results of these statistics are presented in the form of tables and figures. The analysis was done on both the monthly and annual data. These tables and figures show that at the higher fade levels, the number of fade events and the fade time is smaller than at the lower thresholds. For the same fade level the number of fade events and the fade time goes down as the fade duration which it exceeds is increased. Inter-fade durations also showed similar results. The fades exhibited seasonal dependencies. The number of fades (and consequently the fade time) were much higher for the months of May through August and for the months of March and December. The other months showed very little fade activity. A model was also constructed that can predict the fade time as a function of frequency, attenuation level, and fade duration interval. The predicted fade times agree well with the measured fade duration data. An alternate simplified version of the model is also presented. / Master of Science

Frequency scaling of attenuation is the prediction of attenuation at a desired frequency from attenuation values at a base frequency. The attenuation at the base frequency is often known from prior measurements. Frequency scaling of attenuation is of interest because of the eventual need to exploit higher frequency bands. Most satellite communications traffic now use C-band (4-8 GHz) and Ku-band (12-18 GHz). The next approved, yet largely unused, frequency allocation for domestic use is in the K-band (18-27 GHz) to Ka-band (27-40 GHz). At these higher frequencies, however, earth-space radio links suffer atmospherically induced impairments as frequency increases. In particular, rain causes severe fading. Consequently, satellite systems in these higher bands are very susceptible to outages due to rain-induced fades. Reliable frequency scaling models are needed in system design to estimate the effect of these rain-induced fades. Between August 1990 and August 1992, V.P.I. & S.U.'s SATCOM Group performed extensive measurements of slant path attenuation using the 12, 20, and 30 GHz beacon signals (in the Ku/Ka-bands) of the European OLYMPUS experimental satellite. The experimental results are used to evaluate the usefulness of scaling models proposed by other researchers, both for instantaneous and statistical purposes. New models are presented for accurate scaling of attenuation within the Ku/Ka-bands. / Master of Science

M.Ing. (Electrical and Electronic Engineering) / With all the talk about the Information Highway and its construction, there is also a channel which is highly underestimated and thus almost ignored. On normal FM radio transmissions extra bandwidth exists, suitable for the transmission of audio and data. In this thesis the effects of interference on data transmission over the Supplementary Signal Transmission (SST) channel are analysed. The channel is characterized in terms of the Bit Error Rate (BER) versus field strength and distance from a transmitter ...

Interference and jamming severely disrupt our ability to communicate by decreasing the effective signal-to-noise ratio and by making parameter estimation difficult at the receiver. The objective of this research work is to design robust wireless systems and algorithms to suppress the adverse effects of non-intentional co-channel interference (CCI) or intentional jamming. In particular, we develop chip-combining schemes with timing, channel, and noise-power estimation techniques, all of which mitigate CCI or jamming. We also exploit the spatial diversity and iterative receiver techniques for this purpose. Most of the existing timing estimation algorithms are robust against either large frequency offsets or CCI, but not against both at the same time. Hence, we develop a new frame boundary estimation method that is robust in the presence of severe co-channel interference and large carrier-frequency offsets. To solve the high peak-to-average-power ratio problem of a multicarrier code division multiple access (MC-CDMA) system and enhance its robustness against fading and jamming, we propose a constant-envelope MC-CDMA system employing cyclic delay diversity (CDD) as transmit diversity. We analyze the diversity order, coding gain, and bit-error rate upper bound. We also propose a blind, accurate, and computationally efficient signal-to-noise ratio estimator for the proposed system. We propose a configurable robust anti-jam receiver that estimates the frequency- or time-domain jammer state information (JSI) and uses it for chip combining in the corresponding domain. A soft-JSI-based chip-combining technique is proposed that outperforms conventional hard-JSI-based chip combining. We also derive a chip combiner that provides sufficient statistics to the decoder. Channel estimation is necessary for coherent signal detection and JSI estimation. Conversely, knowledge of the jamming signal power and JSI of different subcarriers can improve the accuracy of the channel estimates. Hence, we propose joint iterative estimation of the multiple-input multiple-output (MIMO) channel coefficients, jamming power, and JSI for a coded MC-CDMA MIMO system operating under jamming and a time-varying frequency-selective fading channel. Finally, we reduce the computational complexity of the JSI-based anti-jam receivers by introducing an expectation-maximization-based joint channel and noise-covariance estimator that does not need either the subcarrier JSI or the individual powers of the AWGN and jamming signal.

Thesis (M.App.Sc.) - Carleton University, 2007. / Includes bibliographical references (p. 112-118). Also available in electronic format on the Internet.

Este estudo consiste na comparação de 2 técnicas de modelagem da área de identificação de sistemas aplicadas à predição de dois importantes parâmetros elétricos de um cabo de telecomunicação do tipo LAN, que são Atenuação e Resíduo de Telediafônia. O foco do trabalho é representar a relação existente entre duas maneiras distintas de efetuar a medição dessas grandezas em um cabo. O estudo se iniciou através da aplicação de técnicas de modelagem que buscam Modelos Paramétricos, através da comparação polinomial, baseadas nas respostas dos modelos para os critérios de Akaike, Bestfits e Análise dos resíduos. Na seqüência, a fim de investigar o comportamento não-linear dos parâmetros do cabo em estudo, desenvolveram-se modelos baseados em Redes Neurais Artificiais. Estas redes são técnicas de inteligência artificial, inspiradas na natureza e capazes de identificar padrões e inferir conhecimentos, realizando comparações entre a saída real e a gerada pelo modelo. Utilizaram-se os toolboxes de identificação de sistemas e de redes neurais disponíveis no software Matlab, para desenvolvimento dos modelos. Com estes, neste trabalho foram apresentados os principais conceitos envolvidos em modelagem utilizando modelos paramétricos e redes neurais artificiais. Por fim, são apresentadas comparações entre os resultados obtidos com os respectivos critérios adotados e as técnicas utilizadas apontando para suas virtudes e limitações na solução deste problema. Essas comparações são feitas baseando-se em resultados obtidos para os modelos propostos a partir de dados reais obtidos em testes de qualidade de cabos LAN. / This study consists of a comparison between two system identification modeling techniques applied to predict two important performance parameter involved on a LAN telecommunication cable, which are Insertion Loss and Attenuation Crosstalk Ratio Far End. This work focused on representing the relation between two different ways of measuring a cable by modeling techniques using Parametric Models, which are based on techniques of polynomial comparison performing output models comparison considering Akaike, Bestfit and Residuals analyses criteria and in sequence the Artificial Neural Network, which is an artificial intelligence technique inspired on nature that can identify standards and knowledge, performing a comparison between the real output and the model output. All of that using the Matlab system identification and neural networks toolboxes. On this work were presented the main concepts involved on modeling using parametric models and Neural Networks and was performed a comparison between the obtained results with the adopted criteria and the applied techniques pointing to each advantage and disadvantage for the solution of this problem.

Le besoin incessant de débits toujours plus élevés dans les systèmes de télécommunications a un impact sur tous les éléments composant la chaine de transmission. Ainsi, pour faire face à l’augmentation croissante du volume de données échangées à travers le monde, le développement de nouvelles sources optiques semi-conductrices est absolument nécessaire. La modulation directe de lasers nanostructurés constitue une alternative bas coût et à faible consommation énergétique qui permettra de remplacer progressivement les diodes lasers à puits quantiques actuelles. De nombreux efforts en recherche ont été consacrés au cours des dernières années en vue d’améliorer les performances dynamiques des lasers nanostructurés notamment en terme de bande passante, de facteur de couplage phase-amplitude (facteur α) et de dérive de fréquence (chirp). Pour les applications aux très grands réseaux et systèmes de communication, la croissance d’îlots ou de fils quantiques déposés sur substrat InP permet de réaliser des dispositifs nanostructurés émettant dans le proche infra-rouge autours de 1550 nm. Dans ce mémoire, la dynamique de modulation des lasers nanostructuré est étudiée en régime de modulation directe. Les caractéristiques analysées comprennent: la modulation en amplitude (AM) et en fréquence (FM), le chirp, et les réponses en régime grandsignal. Grâce à une approche semi-analytique, il est démontré que la bande passante et l’amortissement sont fortement limités par les phénomènes de capture et de relaxation des porteurs de charge dans les nanostructures. Afin d’étudier les propriétés du facteur α et du chirp, un nouveau modèle dynamique a été proposé, prenant en compte la contribution à l’indice optique des porteurs de charge dans des états hors résonance. Il est ainsi montré que, contrairement au cas des lasers à puits quantiques, le facteur α dépend fortement du courant de pompe et de la fréquence de modulation. Le facteur α reste constant à basses fréquences (<0,1 GHz) et supérieur aux valeurs obtenues à hautes fréquences (au-delà de quelques GHz) à partir de la technique FM/AM. Ces caractéristiques sont essentiellement attribuées aux contributions des porteurs dans les états hors résonance. Les simulations montrent que le facteur α peut être réduit en augmentant la séparation énergétique entre l’état fondamental résonant (GS) et les états hors résonance. En particulier, un effet laser sur 1’état excité des nanostructures (ES) constitue une solution prometteuse pour améliorer les performances dynamiques, en accroissant notamment la bande passante de modulation et en réduisant le facteur α d’environ 40%. Les techniques d’injection optique sont également intéressantes pour régénérer les performances dynamiques des lasers. Le couplage phase-amplitude et le gain optique y sont substantiellement modifiés via le contrôle de l’amplitude et du désaccord en fréquence du faisceau injecté. Dans ce cadre, ce travail propose une nouvelle technique dérivée de la méthode Hakki-Paoli, permettant de mesurer, sous injection optique, le facteur α à la fois en dessous et au-dessus du seuil. Les lasers à cascade quantique (QCL) sont basés sur des transitions électroniques inter-sous-bandes dans des hétérostructures à puits quantiques. Ces lasers présentent une bande passante (AM) relativement de quelques dizaines de GHz et sans résonance ce qui est prometteur pour les transmissions en espace libre. De manière surprenante, les calculs montrent que les QCL présentent une largeur de bande FM extrêmement large de l’ordre quelques dizaines de THz, environ trois ordres de grandeur supérieure à la largeur de bande AM. L’injection optique dans ces lasers présente les mêmes avantages que ceux procurés dans les lasers à transitions interbandes. Des désaccords positifs ou négatifs en fréquence augmentent notamment la largeur de la bande passante. / High performance semiconductor lasers are strongly demanded in the rapidly increasing optical communication networks. Low dimensional nanostructure lasers are expected to be substitutes of their quantum well (Qwell) counterparts in the next-generation of energy-saving and high-bandwidth telecommunication optical links. Many efforts have been devoted during the past years to achieve nanostructure lasers with broad modulation bandwidth, low frequency chirp, and reduced linewidth enhancement factor. Particularly, 1.55-μm InP-based quantum dash (Qdash)/dot (Qdot) lasers are preferable for long-haul transmissions in contrast to the 1.3-μm laser sources. In this dissertation, we investigate the dynamic characteristics of InPbased nanostructure semiconductor lasers operating under direct current modulation, including the amplitude (AM) and frequency (FM) modulation responses, the linewidth enhancement factor (also known as α-factor), as well as large-signal modulation responses. Using a semi-analytical analysis of the rate equation model, it is found that the modulation bandwidth of the quantum dot laser is strongly limited by the finite carrier capture and relaxation rates. In order to study the α- factor and chirp properties of the quantum dot laser, we develop an improved rate equation model, which takes into account the contribution of carrier populations in off-resonant states to the refractive index change. It is demonstrated that the α-factor of quantum dot lasers is strongly dependent on the pump current as well as the modulation frequency, in comparison to the case of Qwell lasers. The α-factor remains constant at low modulation frequencies (<0.1 GHz) and higher than the value derived at high modulation frequencies (beyond several GHz) from the FM/AM technique. These unique features are mostly attributed to the carrier populations in off-resonant states. Further simulations show that the α-factor can be reduced by enlarging the energy separation between the resonant ground state (GS) and off-resonant states. Lasing from the excited state (ES) can be a promising alternative to enhance the laser’s dynamic performance. The laser exhibits a broader modulation response and the α-factor can be reduced by as much as 40%. The optical injection technique is attractive to improve the laser’s dynamical performance, including bandwidth enhancement and chirp reduction. These are demonstrated both theoretically and experimentally. The phase-amplitude coupling property is altered as well in comparison with the free-running laser and the optical gain depends on the injection strength and the frequency detuning. This work proposes a new method derived from the Hakki-Paoli method, enabling to measure the α-factor of semiconductor lasers under optical injection both below and above threshold. In addition, it is demonstrated theoretically that the α-factor in nanostructure lasers exhibits a threshold discontinuity, which is mainly attributed to the unclamped carrier populations in the off-resonant states. It is a fundamental limitation, preventing the reduction of the α-factor towards zero. Quantum cascade (QC) lasers rely on intersubband electronic transitions in multi-quantum well heterostructures. QC lasers show flat broadband AM response (tens of GHz) without resonance, which constitutes promising features for free-space communications. Surprisingly, calculations show that the QC laser exhibits an ultrabroad FM bandwidth on the order of tens of THz, about three orders of magnitude larger than the AM bandwidth. Optically injection-locked QC lasers also exhibit specific characteristics by comparison to interband semiconductor lasers. Both positive and negative frequency detunings enhance the modulation bandwidth.