Random Data Patterns Research Articles

Predicting the behaviour of G-RAM networks

A Generalising Random Access Memory (G-RAM) neuron is distinguished from conventional neuron models by the fact that its tolerance to departures in similarity from its training pattern is variable. Details of this are given in this paper as it affects the behaviour a class of digital probabilistic neural networks which have been achieving attention in the neural networks literature for some years now. Such systems are also called n-tuple systems, weightless systems or p-RAM systems. After reviewing the literature on such networks, a novel simple combinatoric analysis of the most likely behaviour of recursive GRAM networks is described. The best network performance, measured by a key parameter called ‘radius of retrievability’ (first defined by Wong and Sherrington [J. Phys. A 22 (1989) 2233] as the error in the input that still allows evolution of the dynamic network to the correct attractor state), is obtained with a training set composed of random data patterns. Increasing the size of the training set reduces this radius of retrievability in a predictable manner. Changing the nature of the training set to non-random patterns also reduces the radius of retrievability to an extent that we show can be estimated from a measure of the diversity of the elements of the training set (we refer to this as the ‘mean intra-set Hamming distance of the training set’). As mentioned earlier the feature of G-RAMs (indicated by the G) is that there exists a generalization parameter which determines how far a neuron input vector can stray from a training input for the neuron to respond in the trained way. It is shown that when this generalization parameter is reduced, then the radius of retrievability is also reduced but it is then stable in the face of an increase in size, or change in nature, of the training set. This is a novel prediction of the behaviour of systems and of the robustness of such behaviour in the face of varying the size and correlation properties of the training set.

LVDS I/O interface for Gb/s-per-pin operation in 0.35-μm CMOS

This paper presents the design and the implementation of input/output (I/O) interface circuits for Gb/s-per-pin operation, fully compatible with low-voltage differential signaling (LVDS) standard. Due to the differential transmission technique and the low voltage swing, LVDS allows high transmission speeds and low power consumption at the same time. In the proposed transmitter, the required tolerance on the dc output levels was achieved over process, temperature, and supply voltage variations with neither external components nor trimming procedures, by means of a closed-loop control circuit and an internal voltage reference. The proposed receiver implements a dual-gain-stage folded-cascode architecture which allows a 1.2-Gb/s transmission speed with the minimum common-mode and differential voltage at the input. The circuits were implemented in a 3.3-V 0.35-/spl mu/m CMOS technology in a couple of test chips. Transmission operations up to 1.2 Gb/s with random data patterns and up to 2 Gb/s in asynchronous mode were demonstrated. The transmitter and receiver pad cells exhibit a power consumption of 43 and 33 mW, respectively.

Duration of search for a fixed pattern in randomdata:Distribution function and variance

The distribution function and variance for the duration of search to find a fixed L-ary sequence in a stream of random L-ary equiprobable data are derived. The expressions obtained may be used to evaluate some of the parameters of frame synchronisation in digital communication systems.

A CMOS serial link for fully duplexed data communication

This paper describes a CMOS serial link allowing fully duplexed 500 Mbaud serial data communication. The CMOS serial link is a robust and low-cost solution to high data rate requirements. A central charge pump PLL for generating multiphase clocks for oversampling is shared by several serial link channels. Fully duplexed serial data communication is realized in the bidirectional bridge by separating incoming data from the mixed signal on the cable end. The digital PLL accomplishes process-independent data recovery by using a low-ratio oversampling, a majority voting, and a parallel data recovery scheme. Mostly, digital approach could extend its bandwidth further with scaled CMOS technology. A single channel serial link and a charge pump PLL are integrated in a test chip using 1.2 /spl mu/m CMOS process technology. The test chip confirms upto 500 Mbaud unidirectional mode operation and 320 Mbaud fully duplexed mode operation with pseudo random data patterns. >

A monolithic 156 Mb/s clock and data recovery PLL circuit using the sample-and-hold technique

The PLL circuit described here performs the function of data and clock recovery for random data patterns by using a sample-and-hold technique, and four component circuits (a phase comparator, a delay circuit, a voltage-controlled oscillator, and a S/H switch with a low-pass-filter) were specially designed to further stabilize the PLL operation. A test chip fabricated using Si bipolar process technology demonstrated error-free operation with an input of 2/sup 23/-1 PRBS data at 156 Mb/s. The rms data pattern jitter was reduced to only 1.2 degrees with only an external power supply bypass capacitor.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Frame Synchronization for Gaussian Channels

The problem of locating a periodically inserted frame synchronization pattern in random data for a <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">M</tex> -ary digital communication system operating over the additive white Gaussian noise channel is considered. The optimum maximum-likelihood decision rule, high signal-to-noise approximate maximum likelihood decision rule, and ordinary correlation decision rule for frame synchronization are derived for both coherent and noncoherent phase demodulation. A general lower bound on synchronization probability is derived for the coherent correlation rule. Monte Carlo computer simulations of all three decision rules, along with evaluations of the lower bound for the coherent correlation rule, were performed for the coherent MPSK, coherent, and noncoherent <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">M</tex> ary orthogonal, and 16 QAM signaling schemes. These results show that in each case the high signal-to-noise maximum-likelihood rules have a performance nearly equal to that of the maximum-likelihood rules over a wide range of practically interesting signal-to-noise ratios (SNR's). These high SNR decision rules also provide significant performance improvement over the simple correlation rules. Moreover, they are much simpler to implement than the maximum-likelihood decision rules and, in fact, are no more complex than the correlation rules.

Frame Synchronization for Direct-Detection Optical Communication Systems

The problem of locating a periodically inserted frame synchronization pattern in random data is considered for a directdetection optical communication channel. Both pulse-position modulation and on-off keying with known symbol synchronization are considered. The high signal-to-noise ratio approximation of the maximum likelihood decision rule is derived for each of these cases. Analytical bounds and simulation results on synchronization probability show that the high SNR ML rule has a performance approaching that of the optimum ML rule over a wide range of signal-to-noise ratios. This decision rule also provides a substantial performance improvement over the correlator rule with virtually no additional implementation complexity.

Optical spectra and the system performance of 1.55 μm DFB lasers

Static and dynamic optical spectra of 1.55 μm DCPBH-DFB lasers are analysed. These lasers were packaged in hermetic sealed enclosures and directly modulated with transmitter circuitry. We have found that their static spectra observed under CW operation have fine structure induced by optical feedback from the far end of the package fibre pigtail. The dynamic spectra taken under pulsed modulation differ significantly from the static spectra in two ways. The dynamic spectra are narrower than the structured static spectra when a fixed data pattern is applied to the lasers and broader when a random data pattern is used. This difference is attributed to the finite build-up time required to produce a sustained amplitude oscillation at the relaxation oscillation frequency, within the external optical cavity, after the lasers are turned on. The structured CW and the altered dynamic spectra are found not to degrade the transmission through 50 km of fibre at bit rates as high as 650 Mbit/s.

Modulation Transfer Noise Effects from a Continuous Digital Carrier to FDM/FM Carriers in Memoryless Nonlinear Devices

An analysis of the modulation transfer noise effects from a continuous (nonbursty) digital carrier to multiple FDM/FM carriers in a common memoryless nonlinear amplifier (e.g., satellite transponder TWTA) is presented. It is shown that the modulation transfer noise, which is caused by the random envelope variations of the digital carrier due to the random data patterns and filtering of the PSK pulses, consists of both discrete and continuous spectral components in the baseband of the FDM/FM carriers. Numerical examples in terms of the noise power ratio as well as a comparison between measured and computed results are given. In general, the discrete or tonal interference component at the symbol rate of the digital carrier is quite significant. Results of this paper should be useful in the planning of FDMA satellite transponders employing the mixed PSK-FM mode of operation.

On the expected duration of a search for a fixed pattern in random data (Corresp.)

An expression is obtained for the expected duration of a search to find a given <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">L</tex> -ary sequence in a semi-infinite stream of random <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">L</tex> -ary data. The search time is found to be an increasing function of the lengths of the "bifices" of the pattern, where the term bifix denotes a sequence which is both a prefix and a suffix.