A Design of Input-Decimaiton Technique for Recursive DFT/IDFT Algorithm

In recent years, the OFDM (Orthogonal Frequency Division Multiplexing) modulation method is widely used in high-speed communication systems such as terrestrial digital broadcasting systems and high-speed wireless communication systems. However, the amplitude of the OFDM signal shows large fluctuation. Then an OFDM system requires high linearity, especially in its high power amplifier (HPA). As a result of HPA back-off operation, its efficiency deteriorates, and then its power consumption increases. In this paper, we propose a spatially superposed OFDM transmission system with multiple amplifiers and a spatial power combining technology. We carried out the simulation of the total power consumption and transmission characteristics. From this investigation, it is expected that the proposed system can reduce the power consumption by about 44%, compared with a conventional system with one HPA.

This thesis presents a green design of fast recursive discrete Fourier transform (RDFT) architecture. The advantages of the proposed design include high-throughput, power-efficient, memory-free and more reconfigurable. This thesis also discuss the issue of applying this RDFT design to Modified Discrete Cosine Transform (MDCT) and Inverse Modified Discrete Cosine Transform (IMDCT), which results in an improvement of hardware-sharing and achieves lower-cost. To improve the lack-of-efficiency problem in traditional RDFT architecture, the 2-D architecture algorithm is used in our design. Compared with the latest RDFT architecture proposed by Lai et al, 49.53% reduction of the overall clock cycles is presented in our design; and the computational amount of multiplication and addition presented are reduced to 53.60% and 53.03% of Lai’s design. In addition, the proposed architecture holds four times of DTPT (Data Throughput per Transformation, DTPT) than the traditional architecture; which achieves our goal of presenting a low complexity and low power-cost design. Furthermore, this architecture is applied on Digital Radio Mondiale (DRM) system in our design. Not only DFT computation(288、256、176、112 point) but MPGE-4 HE-AAC techniques are applied on DRM audio coding(1920、240 point). Therefore, to design a common architecture of DFT and IMDCT becomes an important issue for portable DRM receiver. This system is used to examine the performance of the RDFT architecture which we proposed in the front text as well as to discuss other architectures in this article. Moreover, this RDFT processor is implemented by using TSMC 0.18µm 1P6M CMOS technology. The core size is 0.84x0.84mm2 and the power consummation is 14.6mW@25MHz. To conclude, the proposed design would be more efficient and more suitable for DRM applications than previous works.

By using adiabatic perturbation theory, we calculate the timing jitter generated by fluctuations in soliton amplitude, frequency, and position that are induced by the amplifiers noise in high-speed soliton communication systems. The analysis is applied to dispersion-tailored fibers which, in contrast with conventional constant-dispersion fibers, allow ultrashort solitons (width <10 ps) to propagate over long distances with minimum production of dispersive waves. We show that a transition from a regime in which amplifier noise-induced frequency fluctuations dominate the timing jitter (Gordon–Haus jitter) to a regime in which amplitude fluctuations dominate the timing jitter occurs when solitons become shorter than 2–7 ps, depending on the total distance of transmission. The latter source of jitter arises because fluctuations in the soliton amplitude are converted to frequency fluctuations by the Raman effect, which in turn are converted to position fluctuations by the group-velocity dispersion. The contribution of third-order dispersion to the timing jitter is evaluated and discussed. We provide an estimate of the distance at which soliton-control elements (such as synchronous modulators) should be inserted to reduce the timing jitter and extend the transmission distance.

This paper presents a compact structure of recursive discrete Fourier transform (RDFT) with prime factor (PF) and common factor (CF) algorithms to calculate variable-length DFT coefficients. Low-power optimizations in VLSI implementation are applied to the proposed RDFT design. In the algorithm, for 256-point DFT computation, the results show that the proposed method greatly reduces the number of multiplications/additions/computational cycles by 97.40/94.31/46.50% compared to a recent approach. In chip realization, the core size and chip size are, respectively, 0.84 × 0.84 and 1.38 × 1.38 mm2. The power consumption for the 288- and 256-point DFT computations are, respectively, 10.2 (or 0.1051) and 11.5 (or 0.1176) mW at 25 (or 0.273) MHz simulated by NanoSim. It would be more efficient and more suitable than previous works for DRM and DRM+ applications.

GaAs optoelectronic integrated circuits are having an increasing impact on the development of lightwave communication systems and very high sped digital and analog signal processors. However, cost-effective integration of complex electronic and optoelectronic circuits used in high speed networks or backplane interconnects still presents a significant technological challenge and substantial efforts are being devoted to the development of practical, low-loss integration processes. Photodetector/waveguide integration is a key aspect of the successful packaging of optoelectronic devices. In this work, metal-semiconductor- metal (MSM) photodetectors were integrated with polyimide ridge waveguides, and the processing parameters were varied for optimum performance. The absolute responsivities of the integrated MSM/polyimide waveguide structures were typically 0.5 A/W, and 3 dB bandwidths of 4-6 GHz were measured. Series of 1 X 2 and 1 X 4 photodetector arrays were interconnected, demonstrating a uniform division of the optical signal between the detectors. Results indicate that these arrays offer the potential to fabricate optoelectronic switches that can be used in a variety of high speed and broadband communication systems.© (1997) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

"The clock generator is a critical component in high-speed wireline communication systems. The clocks in these systems mark time at precise regular intervals. Any deviation of the clock edges from these ideal time points is called jitter and it degrades the overall integrity of the communication system. This thesis looks at both phase locked loops (PLLs) and delay locked loops (DLLs) and their ability to provide these precise clocks. Expressions for the thermal noise-induced jitter in both the PLL and DLL are compared. From this, the DLL is shown to possess the potential for better jitter performance. However, to realise the potential of the DLL, another source of jitter peculiar to the DLL has to be reduced. This jitter in DLLs is caused by Static Phase Offset (SPO). The thesis details the causes of SPO and describes a number of good design practices for reducing it. It then describes an input and feedback clock phase interpolation technique which reduces SPO by half. Finally a novel chopping technique is used to essentially eliminate SPO resulting in a DLL which is ideal for high-speed wireline communication systems. Some of this work was published at the lEE Irish Signals and Systems Conference, which took place at Queens University Belfast, in June 2004."

In this paper, the problem of fast time-varying channel prediction is investigated in high-speed railway communication systems. A channel prediction algorithm is proposed based on a support vector machine (SVM) model. In order to further improve the prediction accuracy, the penalty coefficient and Gaussian kernel width of the SVM model are optimized by a genetic algorithm (GA). Simulation results show that the proposed prediction model based on both the SVM and the GA (SVM-GA) has lower prediction error than traditional auto-regressive (AR) and single SVM prediction models. In addition, when the effect of the noise on prediction performance is considered, the SVM-GA prediction model is superior to the AR and the SVM models in terms of normalized mean squared error.

A new scheme to realize terahertz (THz) radiation generation, as well as high-speed THz modulation, for future high-speed THz communication systems, is proposed and discussed. The proposed experimental scheme includes two elemental parts: dual-wavelength mode-locking oscillating cavity and photomixing elements. The conditions for double-wavelength operation using the proposed experimental setup have been theoretically discussed and obtained. The mode-locked output optical short pulses with two highstability wavelengths are obtained and demonstrated. These can be used to produce THz radiation, as well as high-speed signal modulation, through the photomixing effects in an ultrafast oxygen-ion-implanted and epitaxial InGaAs:O for the future high-speed THz communication systems.

The two-section gain-coupled DFB laser can be designed to exhibit self pulsing. The self-pulsing dynamics have been exploited to demonstrate a variety of application in high-speed optical communication systems, including wavelength and polarization-independent all-optical clock recovery up to 100 Gb/s and conversion from the NRZ or RZ format to CSRZ format.

FH(frequency hopping) is widely concerned in the field of communication because of its outstanding ability of noise suppression. The suppression ability of FH communication can be improved by increasing the bandwidth and hopping rate. MSK is chosen as a modulation of high-speed broadband FH communication system to research the suppression ability of FH to different noise signals. The diagram between ratio peak and ISR (interference-signal power) can be given in different bandwidths and different hopping rates. It is verified that the broadband high-speed FH communication system with MSK as the modulation mode has not only a smaller BER (bit error rate) but also a higher noise suppression.

In this paper, optimized transmit schemes for multiple-input multiple-output (MIMO) systems with simplified receivers are proposed for the downlink of high-speed wireless communication systems. In these systems, MIMO signal preprocessing is performed at the transmitter or base station with the receiver at the mobile station having a simplified structure that requires only limited signal processing. An important potential application for our transmit MIMO techniques is in the downlink of high-speed wireless communication systems with Vertical Bell Laboratories Layered Space-Time (V-BLAST) or a similar technique utilized in the uplink, creating a high-speed duplex system with a simplified mobile station transceiver structure. Two approaches are introduced and these depend on whether or not receive diversity is employed at the receiver. Both methods require that channel state information be available at the transmitter. In addition, some important associated issues such as peak-to-average power ratio requirements at the transmitter and robustness to downlink channel errors are also investigated and various solutions are proposed. Simulation results are provided and these show that performance improvement can be achieved when compared with other MIMO transmit schemes.

Wide-band digital indoor and future mobile communication systems are being developed in the quasi-millimeter and millimeter wave frequency bands. In such high-speed systems, the degradation of transmission quality by multipath fading constitutes a limiting barrier. Multipath fading, which results from reflection, refraction and scattering of radio waves by buildings and other structures, gives rise to more than one path reaching the receiver and produces a distorted version of the transmitted signal. In mobile and indoor communication systems it cannot be eliminated, therefore the multipath channel must be well characterized in order to reduce its effect in the design of such systems. The objective of this research is to study the multipath fading and delay spread effects due to buildings in an urban propagation channel. In particular, the effects of the building materials, shapes and number of buildings on multipath fading are being studied.

This paper presents low latency and/or computation algorithms of iterative codes of turbo codes, turbo product codes and low density parity check codes for use in wireless broadband communication systems. Due to high coding complexity of iterative codes, this paper focus on lower complexity and/or latency algorithms that are easily implementable in hardware and further accelerate the decoding speed.

Providing reliable and quality of service (QoS) guarantee service is basic requirement for communication system. However, the intrinsic features of high-speed railway communication (HSRC), such as serious channel distortion, lack of perfect channel state information (CSI), severe Doppler effect, and sensitive to algorithm computation complexity, make this requirement hard to be met. In this paper, considering the intrinsic features of HSRC, a novel link adaptation scheme is proposed to improve HSRC systems goodput while guaranteeing prescribed error rate target. We formulate the link adaptation as a partially observable Markov decision process (POMDP). With this stochastic optimization formulation, the proposed link adaptation scheme can tackle the imperfect CSI and has very low computation complexity. In addition, cross-layer design methodology is exploited, where diversity/multiplexing gain selection in physical layer and frame size in link layer are considered together to improve system goodput. Simulation results demonstrate the effectiveness of the proposed link adaptation scheme.

Providing high-speed communication for mobile users in remote geographic areas or after a disaster occurs is not only critical but also challenging. To counter such challenge, unmanned aerial vehicles (UAVs) have been exploited to provide a fast-deployable and high-speed communication system, where each UAV can serve as an aerial small cell base station to provide WiFi and/or cellular services for the ground users. Despite its fast-deployable and highly maneuverable features, the capacity analysis of aerial small cells is largely missing. To close such gap, a stochastic propagation model for A-to-G aerial channels is first introduced, which takes into account the impact from wave propagation, gaseous absorption, Doppler spread, attitude-dependent shadowing, and multipath fading. Then, by exploiting such model, the area spectral efficiency of the aerial small cells is investigated for both SISO and MIMO cases. Our study reveals the inherent relationship among the area capacity, height and coverage and shows that there exists an optimal attitude that can maximize network capacity and cell coverage.