Noise-correlating radar based on retrodirective antennas

A new retrodirective antenna-based search radar system has been introduced. The suggested system uses a noise correlation technique to detect the presence and the direction of the target. Simulation and analytical results show an order of magnitude improvement in acquisition time of the radar when compared with a phased array antenna-based radar system with the same specifications, except transmit power. To the best knowledge of the authors, no radar of a comparable acquisition time has been designed to this date. Power versus acquisition time tradeoff has been compared with a phased array radar for evaluating performance of the system. The radar is self-tracking due to retrodirectivity of the antenna array, and is much easier to implement, as it does not require any phase shifters etc.     

Threshold Investigation of Phase-Locked Discriminators

This paper is primarily an investigaton of a second-order phased-locked loop in the threshold region, with the loop filter chosen as (1/1+?s) or (1+?1s/1+?s). First-order loop results are also given. Effects of loop parameters, detuning error, and modulation are studied. It is shown that the loop with filter (1+?1s/1+?s) can be built two ways with same bandwidth and damping. One loop gives much better threshold than the other. The analytical results are supported by computer simulation as well as experimental work obtained by hardware simulation of the loop.     

On Higher Order Discrete Phase-Locked Loops

An exact mathematical model is developed for a discrete loop of a general order particularly suitable for digital computation. The deterministic response of the loop to the phase step and the frequency step is investigated. The design of the digital filter for the second-order loop is considered. Use is made of the incremental phase plane to study the phase error behavior of the loop. The model of the noisy loop is derived and the optimization of the loop filter for minimum mean-square error is considered.     

Statistical Analysis of Digital Phase-Locked Loops Fading Channiels

Statistical analyses of zero-crossing sampling first- and second order digital phase-locked loops in fading channls (DPLFC) are presented. A DPLFC is defined as one where the incoming signal is not only corrupted by additive Gaussian noise channel (AGN) but also by a fading communication channel. The three fading channels considered are the log-normal, Rayleigh, and Rician channels. The performance degradation of the DPLFC operation is characterized by the steady-state phase-error process probability density function and phase-error variance. Approximate analytic expressions for the phase-error statistics of a first-order DPLFC for both phase-step and frequency-step inputs are obtained by solving the Chapman-Kolmogorov equation associated with the phase-error process. The second-order DPLFC is analyzed for a frequency-step input. Numerical and simulation results are provided confirming the analyses presented for the three fading channels considered.     

Mesospheric ionisation over dip equator at sunrise

The electron density profile in the equatorial mesosphere was measured during sunrise time over Thumba(dip lat= 0.6°S). The measurements were carried out in the altitude range 60 to 100 km using rocketborne probes. A sharp layer of ionisation was observed around 80 km with electron density about 10⁸m^?3. It is suggested that hydrated ions are the main constituents of this layer.     

A Modified Karhunen-Loeve Model of Gravity Disturbance Vector

A modified anisotropic Karhunen-Loeve (KL) model for the gravity disturbing potential is proposed. This model, unlike the previous KL model of Bose, does not vanish on the boundaries of a local region. Laplace's equation for the gravity disturbing potential and flatearth Vening-Meinesz relationships of the gravity disturbance vectorcomponents (?,?,?g) when the model is isotropic, are exploited to determine the unknown parameters of the model. The resulting model, in addition to having random boundary values, has three more parameters than those found in Bose's KL model. This should provide extra flexibility in fitting the model to the noisy data available in a local finite region.