Engineering topics
Publications and source records attributed to Rubin, A. L..
Mathematical analysis for digital data transmission calculates optimum number of binary error-correcting repeaters to install in given number of wideband channel links. Asymptotic results compared to computed numerical results.
The present investigation is concerned with the error control strategy used in connection with digital data networks, taking into account the capacity of the end-to-end channel. It is assumed that the only source of noise is Gaussian and that bandwidth limitation on capacity can be ignored. Attention is given to L links in tandem with binary repeaters, the asymptotic capacity with a large number of links, some alternative repeater strategies, and approaches for lowering the bit error probability. Numerical results are presented, taking into consideration capacity versus link error probability for various numbers of links.
A new digital processing algorithm based on the fast polynomial transform is developed for producing images from Synthetic Aperture Radar data. This algorithm enables the computation of the two dimensional cyclic correlation of the raw echo data with the impulse response of a point target, thereby reducing distortions inherent in one dimensional transforms. This SAR processing technique was evaluated on a general-purpose computer and an actual Seasat SAR image was produced. However, regular production runs will require a dedicated facility. It is expected that such a new SAR processing algorithm could provide the basis for a real-time SAR correlator implementation in the Deep Space Network. Previously announced in STAR as N82-11295
Field programmable logic array (FPLA) is used to make 3-bit arithmetic logic unit (ALU) for large integrator that can be read and cleared while new data is added to begin new integral. Arrangement of device can provide for full carry/lookahead capability with minimum gate delays.
A 4-bit and a 3-bit adder are described which can be implemented under special hardware restrictions. The chip to be used is field-programmable logic array (FPLA) with 12 input lines, 50 AND gates inside, and output through only 6 OR gates. The context in which it is being used requires an enable function which can suppress one of the two numbers to be added. The 3-bit enabled adder is compatible with lookahead-carry mechanizations using the 74S182.
Electrophoretic mobilities (EPM) of peripheral lymphocytes were studied from normal subjects, chronic hemodialysis patients and kidney transplant recipients. A technique to separate B lymphocytes and null cells from non-T lymphocyte preparation was developed. The experiments were designed to determine which subpopulation of the non-T lymphocytes is primarily affected and shows a decreased EPM in chronic hemodialysis patients and kidney transplant recipients.
Electrophoretic mobility (EPM) of human peripheral lymphocytes were examined with the following objectives: To determine differences in EPM of lymphocytes under immuno-stimulated and immuno-suppressed states. To define the conditions necessary for the separation of lymphocyte sub-populations in normal and pathological conditions; To investigate immunological active, charged chemical groups on lymphocyte surfaces; and to investigate pathophysiological mechanisms of immune responsiveness, as reflected by alterations in EPM. To evaluate the potential of lymphocyte electrophoresis as: (1) a means of monitoring the immune status of kidney transplant recipients, (2) in predicting the outcome of kidney transplants, and (3) as a method for separation of lymphocyte sub-populations, the EPM was studied for unfractionated human peripheral lymphocytes and of populations enriched with T and "B" cells from normal adults, hemodialysis patients and kidney transplant recipients.
The capacity region of a multiple-access channel has recently been identified as the convex hull (barred K) of a certain set (K) of points in the first quadrant of the (R1,R2) plane. For a pair of rates in K, a more or less standard random-coding argument can be used to show the existence of a good pair of codes. But for points in barred K-K, it is apparently necessary for the two senders to use some form of time sharing to achieve the desired rates. However, in order to share time, at least one of the senders must have knowledge of the other's phase; and in many practical situations this knowledge does not exist. This paper investigates the problems which arise in coding for multiple-access channels when the senders cannot synchronize with each other.