Experiments on real-time 1-Gb/s packet transmission using MLD-based signal detection in MIMO-OFDM broadband radio access

This paper presents experimental results on real-time packet transmission of greater than 1 Gb/s using 4-by-4 multiple-input-multiple-output (MIMO) multiplexing and maximum-likelihood detection (MLD)-based signal detection with a decreased level of computational complexity in orthogonal frequency-division multiplexing (OFDM) radio access. We apply our previous algorithm called adaptive selection of surviving symbol replica candidates (ASESS) based on the maximum reliability in MLD employing QR decomposition and the M-algorithm (QRM-MLD) to reduce the extremely high level of computational complexity in the conventional MLD. The experimental results using multipath fading simulators are in good agreement with the computer simulation results. The loss in the required received signal energy per bit-to-background noise power spectrum density ratio (E/sub b//N/sub 0/) is suppressed to within approximately 1-2 dB. Therefore, through experiments, we demonstrate that the QRM-MLD employing ASESS is very beneficial in reducing the influence of hardware implementation loss, as well as in decreasing the required received E/sub b//N/sub 0/. We further show that the extremely high-speed real-time packet transmission of greater than 1 Gb/s in a 100-MHz channel bandwidth (i.e., 10 bit/s/Hz) is achieved at the average received E/sub b//N/sub 0/ per receiver antenna of approximately 12 dB using 16QAM modulation and turbo coding with the coding rate of 8/9 in 4-by-4 MIMO multiplexing.     

Broadband 180/spl deg/ bit phase shifter using a /spl lambda//2 coupled line and parallel /spl lambda//8 stubs

In this paper, a broadband 180/spl deg/ bit phase shifter using a new switched-network was presented. The new network is composed of a /spl lambda//2 coupled line and parallel /spl lambda//8 open and short stubs, which are shunted at the edge points of a coupled line, respectively. According to a desired phase shift, it provides a controllable phase dispersive characteristic by the proper determination of Z/sub m/,Z/sub s/, and R values. The 180/spl deg/ bit phase shifter operated at 3 GHz was fabricated and experimented using design graphs which provide the required Z/sub m/,Z/sub s/ values, and I/O match and phase bandwidths. The experimental performances showed broadband characteristics.     

Equivalent Circuit of the Bolometer Detector

A small-signal dynamic equivalent circuit is established for the output voltage of a dc-biased bolometer (barretter) detector. The circuit consists of a voltage generator /spl upsi//sub g/, whose output is an undistorted replica of the incident RF-power modulation envelope, followed by a series resistor R/sub 1/ of dynamic origin, a shunt capacitor C that represents heat storage in the bolometer wire, and a series resistor R/sub 0/ equal to the dc resistance, usually 200 ohms. The resistance R/sub 1/ is independent of signal level, and is typically about 220 ohms for an 8.75-mA bolometer and about 120 ohms for a 4.5-mA bolometer. At a modulation frequency f/sub m/ near 0 Hz, the equivalent audio source impedance of the bolometer is R/sub 1/ +R/sub 0/. The common belief that the source impedance is R/sub 0/ in the weak-signal case is, therefore, refuted. Formulas are derived giving v/sub g/ / /P/sub RF/ and R/sub 1/ as functions of basic, easily determined bolometer parameters. The time constant for open-circuit load is /spl tau//sub oc/= R/sub 1/C, where /spl tau//sub oc/ is determined best by measurement, since catalog values of /spl tau//sub oc/ often are seriously in error. The capacitance is C=/spl tau//sub oc/ / /R/sub 1/. With one type of bolometer /spl tau//sub oc/ measures about 110 /spl mu/s, while various catalogs state values of 250 to 350 /spl mu/s. The equivalent circuit is confirmed quantitatively by measurements of output voltage and source impedance versus modulation frequency.     

Design-for-testability and fault-tolerant techniques for FFT processors

In this paper, we first propose a novel design-for-testability approach based on M-testability conditions for module-level systolic fast Fourier transform (FFT) arrays. Our M-testability conditions guarantee 100% single-module-fault testability with a minimum number of test patterns. Based on this testable design, fault-tolerant approaches at the bit level and the multiply-subtract-add (MSA) module level are proposed, respectively. If the reconfiguration is performed at the bit level, then the FFT/sub BIT/ network is constructed. Two types of reconfiguration schemes (Type-I FFT/sub MSA/ and Type-II FFT/sub MSA/) are proposed at the MSA module level. Since both the design for testability (DFT) and the design for yield (DFY) issues are considered at the same time for all these proposed approaches, the resulting architectures are simpler as compared with previous works. The reliability of the FFT system increases significantly. The hardware overhead is low-about 12% and 1/2N for the FFT/sub BIT/ network and the Type-II FFT/sub MSA/ network, respectively. An experimental chip is also implemented to verify our approaches. Reliabilities and hardware overhead are also evaluated and compared with previous works.     

A theoretical high-rate analysis of causal versus unitary online transform coding

Backward adaptive or "online" transform coding (TC) of Gaussian sources is investigated. The Karhunen-Loe/spl grave/ve transform (KLT, unitary approach) and the causal transform (CT, causal approach) are compared in this context. When the covariance matrix R_/sub x/ of the source is used in the TC scheme, KLT and CT present similar coding gains at high rates. The aim of this study is to model analytically the behavior of these two coding structures when the ideal TC scheme gets perturbed, that is, when only a perturbed value R_/sub x/+/spl Delta/R is known at the encoder. In the online TC schemes considered here, this estimate is used to compute both the transform and the bit assignment. /spl Delta/R is caused by two noise sources: estimation noise (finite set of available data at the encoder) and quantization noise (quantized data at the decoder). Furthermore, not only does the transformation itself get perturbed, but also the bit assignment does as well. In this framework, theoretical expressions for the coding gains in both the unitary and the causal cases are derived under the high-rate assumption.     

Efficient VLSI digital logarithmic codecs

N bit digital words can be logarithmically encoded and compressed to a word length of (log/sub 2/n+m-1) bit maintaining a relative accuracy of m bit over (n-m) octaves of signal level. A bit-serial VLSI coder is reported, which requires little more than a log/sub 2/n counter and an output register and it has a latency of one wordlength. The bit-parallel coder can be built with less than n/sup 2/ transistors and has less than n/4 gate delays. The decoder has similar properties and it expands the logarithm to an antilogarithm with n bit of dynamic range. Using these codecs, digital multiplication, division, powers and roots are reduced to additions, subtractions and shifts, respectively.<>     

A new single-fiber 10-gb/s optical loopback method using phase Modulation for WDM optical access networks

This paper proposes a new single-fiber 10-Gb/s optical loopback method for wavelength-division-multiplexing (WDM) optical access networks. The proposed method improves the tolerance of backreflection that causes interferometric crosstalk with the loopback method on a single fiber. Furthermore, bit synchronization is used to avoid the problem of wavelength dispersion caused in optical access networks. The authors simulated the available distance, received eye patterns at 10 Gb/s, and confirmed the excellent eye opening. 10-Gb/s back-to-back experiments show that the proposed method provides a 17-dB improvement in backreflection tolerance compared with the direct method. The paper reports the first successful results for 10- and 20-km transmission experiments using a standard single-mode optical fiber. The measured eye patterns agreed well with the simulation results. The paper also shows measured bit-error rates (BERs) equivalent to those of a back-to-back configuration.     

Temperature and substrate-impedance dependence of noise figure of monolithic RF inductors on silicon

In this letter, we analyze the effects of temperature (from -50/spl deg/C to 200/spl deg/C) and substrate impedance on the noise figure (NF) and quality factor (Q-factor) performances of monolithic RF inductors on silicon. The results show a 0.75 dB (from 0.98 to 0.23 dB) reduction in minimum NF (NF/sub min/) at 8 GHz, an 86.1% (from 15.1 to 28.1) increase in maximum Q-factor (Q/sub max/), and a 4.8% (from 16.5 to 17.3 GHz) improvement in self-resonant frequency (f/sub SR/) were obtained if post-process of proton implantation had been done. This means the post-process of proton implantation is effective in improving the NF and Q-factor performances of inductors on silicon mainly due to the reduction of eddy current loss in the silicon substrate. In addition, it was found that NF increases with increasing temperature but show a reverse behavior within a higher frequency range. This phenomenon can be explained by the positive temperature coefficients of the series metal resistance (R/sub s/), the parallel substrate resistances (R/sub sub1/ and R/sub sub2/), and the resistance R/sub s1/ of the substrate transformer loop. The present analyzes are helpful for RF designers to design less temperature-sensitive high-performance fully on-chip low-noise-amplifiers (LNAs) and voltage-controlled-oscillators (VCOs) for single-chip receiver front-end applications.     

A high-speed 850-nm optical receiver front-end in 0.18-/spl mu/m CMOS

A high-speed optical interface circuit for 850-nm optical communication is presented. Photodetector, transimpedance amplifier (TIA), and post-amplifier are integrated in a standard 0.18-/spl mu/m 1.8-V CMOS technology. To eliminate the slow substrate carriers, a differential n-well diode topology is used. Device simulations clarify the speed advantage of the proposed diode topology compared to other topologies, but also demonstrate the speed-responsivity tradeoff. Due to the lower responsivity, a very sensitive transimpedance amplifier is needed. At 500 Mb/s, an input power of -8 dBm is sufficient to have a bit error rate of 3/spl middot/10/sup -10/. Next, the design of a broadband post-amplifier is discussed. The small-signal frequency dependent gain of the traditional and modified Cherry-Hooper stage is analyzed. To achieve broadband operation in the output buffer, so-called "f/sub T/ doublers" are used. For a differential 10 mV/sub pp/ 2/sup 31/-1 pseudo random bit sequence, a bit error rate of 5/spl middot/10/sup -12/ at 3.5 Gb/s has been measured. At lower bit-rates, the bit error rate is even lower: a 1-Gb/s 10-mV/sub pp/ input signal results in a bit error rate of 7/spl middot/10/sup -14/. The TIA consumes 17mW, while the post-amplifier circuit consumes 34 mW.     

Determination of source and drain parasitic resistances of HEMTs

A practical method to evaluate the parasitic source resistance R/sub s/ and drain resistance R/sub d/ of high electron mobility transistors (HEMTs) is presented. DC results for R/sub s/ and R/sub d/ can be used as initial values in the determination of linear and nonlinear equivalent circuits by optimisation.<>     

Wyner-Ziv coding based on TCQ and LDPC codes

This paper considers trellis coded quantization (TCQ) and low-density parity-check (LDPC) codes for the quadratic Gaussian Wyner-Ziv coding problem. After TCQ of the source X, LDPC codes are used to implement Slepian-Wolf coding of the quantized source Q(X) with side information Y at the decoder. Assuming 256-state TCQ and ideal Slepian- Wolf coding in the sense of achieving the theoretical limit H(Q(X)|Y ), we experimentally show that Slepian-Wolf coded TCQ performs 0.2 dB away from the Wyner-Ziv distortionrate function D/sub WZ/R) at high rate. This result mirrors that of entropy-constrained TCQ in classic source coding of Gaussian sources. Furthermore, using 8,192-state TCQ and assuming ideal Slepian-Wolf coding, our simulations show that Slepian-Wolf coded TCQ performs only 0.1 dB away from D/sub WZ/R) at high rate. These results establish the practical performance limit of Slepian-Wolf coded TCQ for quadratic Gaussian Wyner-Ziv coding. Practical designs give performance very close to the theoretical limit. For example, with 8,192-state TCQ, irregular LDPC codes for Slepian-Wolf coding and optimal non-linear estimation at the decoder, our performance gap to DWZ(R) is 0.20 dB, 0.22 dB, 0.30 dB, and 0.93 dB at 3.83 bit per sample (b/s), 1.83 b/s, 1.53 b/s, and 1.05 b/s, respectively. When 256-state 4-D trellis-coded vector quantization instead of TCQ is employed, the performance gap to DWZ(R) is 0.51 dB, 0.51 dB, 0.54 dB, and 0.80 dB at 2.04 b/s, 1.38 b/s, 1.0 b/s, and 0.5 b/s, respectively.     

Rate-distortion approach to databases: storage and content-based retrieval

This paper investigates the relationship between rate-distortion theory and efficient content-based data retrieval from high-dimensional databases. We consider database design as the encoding of a data object sequence, and retrieval from the database as the decoding of the sequence using side information (i.e., the query) available only at the decoder. We show that, in this setting, the optimal asymptotic tradeoff between the search time R/sub s/ (bits per data object read from the storage device) and the expected search accuracy D/sub s/ (relevance of the retrieved data set) is given by the Wyner-Ziv solution with a side-information-dependent distortion measure. Moreover, the data indexing and retrieval problem is, in general, inseparable from the data compression problem. Data items selected by the search procedure, which can be stored in the disk with a limited total rate of R/sub r/ /spl ges/ R/sub s/, need to be presented at a prescribed expected reconstruction quality D/sub r/. This is, hence, a problem of scalable source coding or successive refinement, albeit with differing layer distortion measures to quantify search and reconstruction quality, respectively. We derive a single-letter characterization of all achievable quadruples {R/sub s/,R/sub r/,D/sub s/,D/sub r/}, and prove conditions for "successive refinability" without rate loss. Finally, we show that the special case D/sub s/=D/sub r/=0 is nontrivial and of practical interest in this context, as it can impose "acceptable" search and reconstruction qualities for each individual data item and for the entire query space with high probability, in contradistinction with standard average distortion requirements. The region of achievable {R/sub s/,R/sub r/} is obtained by adapting Rimoldi's characterization to a new regular scalable coding problem.     

Postcanceling techniques for simultaneous broadcasting of analog FM and digital data

We consider simultaneous broadcasting of low-power digital data and analog FM and present reliable receivers for the digital data. Due to the relatively low power level of the digital data and the interference suppression capability of analog FM, standard analog FM receivers can reliably recover the analog FM audio signal. To recover the digital data, an extended Kalman filter front end is developed that exploits the capture capability of analog FM to reconstruct and postcancel the analog FM component from the received composite signal. Simulations are conducted with artificial analog FM signals, suggesting that postcanceling schemes can provide higher data rates than their precanceling counterparts, at a lower transmission delay penalty but higher digital receiver complexity. For analog FM to digital signal power level ratios in the range of 30-50 dB, the postcanceler digital signal recovery appears fairly robust, providing digital signal-to-noise ratios of 2-7.5 dB. The corresponding uncoded bit error rates strongly depend on the power level difference between the host analog FM and the digital data signal. In particular, at 260 kb/s and E/sub b//N/sub o/=10 dB they range between 1% to about 15%, and can be reduced to acceptable levels using standard channel coding techniques.     

100-Gb/s 2/sup 7/-1 and 54-Gb/s 2/sup 11/-1 PRBS generators in SiGe bipolar technology

This paper presents two monolithic pseudorandom bit sequence (PRBS) generators. One circuit uses a seven-stage shift register operating with a half-rate clock and provides output signals up to 100 Gb/s. The second circuit contains an eleven-stage shift register operating with a full-rate clock up to 54 Gb/s. Both PRBS generators provide a wide range of data rates down to below 1 Gb/s simply by changing the frequency of the external clock signal without the need of any further adjustments. The integrated circuits provide a trigger output which can be switched between eye and pattern display. Furthermore, they contain additional circuitry to guarantee automatic start after power-on. The circuits are manufactured in a 200-GHz f/sub T/ SiGe bipolar technology. They each have a chip size of 900/spl times/700 /spl mu/m/sup 2/ and consume 1.5 and 1.9 W, respectively.     

Microwave Measurement of Dielectric Properties of Low-Loss Materials by the Dielectric Rod Resonator Method

Improvements both in accuracy and speed are described for the technique of measuring the microwave dielectric properties of low-loss materials by using a dielectric rod resonator short-circuited at both ends by two parallel conducting plates. A technique for measuring the effective surface resistance R/sub s/ of the conducting plates is proposed to allow the accurate measurement of the loss tangent tan delta. By means of the first-order approximation, the expressions are analytically derived for estimating the errors of the measured values of relative permittivity epsilon/sub r/, tan delta, and R/sub r/, for measuring the temperature coefficient of epsilon/sub r/, and for determining the required size of the conducting plates. Computer-aided measurements are realized by using these expressions. It is shown that the temperature dependence of R/sub s/, should be considered in the tan delta measurement. The copper plates used in this experiment have the relative conductivity of 91.0+-2.7 percent at 20°C, estimated from the measured R/sub s/ value. For a 99.9-percent alumina ceramic rod sample, the results measured at 7.69 GHz and 25°C show that epsilon/sub r/,= 9.687+-0.003 and tan delta = (1.6+-0.2)x 10/sup -5/. The temperature coefficients measured between 25 and 100°C are 112x10/sup -6//°C for epsilon/sub r/, and 23x10/sup -4//°C for tan delta.     

A 4 Gb/s 2-level to 2 Gsymbol/s 4-level converter GaAs IC forsemiconductor optical amplifier QPSK modulators

The design of a 50 Ω impedance matched two-to-four level converter GaAs IC for two-electrode semiconductor optical amplifier (SOA) modulators is presented. The designed IC exhibits eye diagrams with eye openings of better than 0.30 V and a spacing between adjacent output signal levels of 0.33 V for output symbol rates of up to 2 Gsymbol/s corresponding to input bit rates of up to 4 Gb/s. A novel differential super buffer output driver is applied, for which output reflection coefficients |S₂₂| of less than -12 dB for frequencies up to 10 GHz are obtained. A 1 Gb/s optical QPSK microwave link transmission experiment using a packaged sample of the designed IC and a two-electrode semiconductor optical amplifier phase modulator has been conducted     

Simultaneous all-optical frequency downconversion technique utilizing an SOA-MZI for WDM radio over fiber (RoF) applications

Simultaneous all-optical frequency-downconversion technique utilizing a semiconductor optical amplifier Mach-Zehnder interferometer (SOA-MZI) is experimentally demonstrated, and its application to a wavelength-division-multiplexing (WDM) radio over fiber (RoF) uplink is proposed. The conversion efficiencies from 22.5 (f/sub RF/) to 2.5 GHz (f/sub IF/=f/sub RF/-2f/sub LO/) are in the range from 1.5 to 3 dB for the optical RF wavelength between 1548 and 1558 nm. Error-free simultaneous all-optical frequency downconversion of the two WDM RoF upstream channels that carry 155-Mb/s differential phase-shift keying data at 22.5 GHz to an optical intermediate frequency signal having the frequency of 2.5 GHz with the power penalty less than 0.1 dB at the bit error rate of 10/sup -8/ is achieved.     

Simple method to determine series resistance and its temperature dependence in AlAs/GaAs/AlAs double barrier resonant tunnelling diodes

The author presents a simple empirical method to quickly determine the series resistance R/sub s/ of AlAs/GaAs/AlAs double barrier resonant tunnelling diodes, and its temperature dependence over the temperature range 198-295 K. At high voltage biases, the quantum well diode current-voltage characteristics were modelled with the empirical expression I/sub D/ approximately= exp (q(V/sub A/-I/sub D/.R/sub s/)/kT). From this current-voltage relationship, R/sub s/ can be extracted from plots of the total resistance dV/sub A//dI/sub D/ against 1/I/sub D/ in the limit of 1/I/sub D/ to 0. The results showed that R/sub s/ decreased with temperature from 5.3 Omega at 198 K to 4.7 Omega at 295 K. R/sub s/ values extracted using the Nelder-Meade simplex algorithm agreed with those from the dV/sub A//dI/sub D/ against 1/I/sub D/ technique to within 10%.<>     

Investigation of maximum current sensing window for two-side operation, four-bit/cell MLC nitride-trapping nonvolatile flash memories

Localized charges in a nitride-trapping device provide two-bit/cell operations. Adding multilevel-cells (MLCs) to the physical bits produces a four-bit/cell device. However, it is difficult to get sufficient sensing windows for MLC operation because the left bit and right bit interfere with each other. This letter analyzes the effect of the second bit effect and investigates parameters affecting the sensing current window for physical four-bit/cell operations. The sensing window is found to increase with a higher reading bias, and also with a higher programmed V/sub t/. However, severe second bit effects set in at high V/sub t/, and decreased the sensing window again. An optimal sensing window is found at moderately high V/sub t/.