Parallel and serial heuristics for the minimum set cover problem

We present a theoretical analysis and an experimental evaluation of four serial heuristics and four parallel heuristics for the minimum set cover problem. The serial heuristics trade off run time with the quality of the solution. The parallel heuristics are derived from one of the serial heuristics. These heuristics show considerable speedup when the number of processors is increased. The quality of the solution computed by the heuristics does not degrade with an increase in the number of processors.Research of both authors was supported by NSF Grant No. MIP-8807540.     

Evaluation of materials for artificial heart valves

The heart of a normal human being beats about 38 million cycles every year. An artificial heart valve, to perform at this rate in the adverse conditions inside the heart for 20 years or more, should be highly wear-resistant with excellent fatigue strength. Thus, the study of mechanical and physical properties of the materials intended for use in artificial valves becomes an inseparable part of the valve development process itself. The physical and mechanical requirements of the materials used in the Chitra heart valve have been evaluated by studying their water absorption, adhesive wear and abrasive properties. The mechanical durability of the device has been assessed by accelerated life cycle testing. The test systems developed for the above are described here. The results show UHMW-PE to be a highly wear-resistant material suitable for the occluder. The accelerated wear tests show that the valve with Haynes-25 alloy cage and UHMW-PE disc has durability in excess of 50 years.     

Efficient Transition Fault ATPG Algorithms Based on Stuck-At Test Vectors

This paper proposes novel algorithms for computing test patterns for transition faults in combinational circuits and fully scanned sequential circuits. The algorithms are based on the principle that s@ vectors can be effectively used to construct good quality transition test sets. Several algorithms are discussed. Experimental results obtained using the new algorithms show that there is a 20% reduction in test set size, test data volume and test application time compared to a state-of-the-art native transition test ATPG tool, without any reduction in fault coverage. Other benefits of our approach, viz. productivity improvement, constraint handling and design data compression are highlighted.     

Algorithms for I DDQ measurement based diagnosis of bridging faults

In the absence of information about the layout one is left with no alternative but to consider all bridging faults. An algorithm for diagnosis of a subset of such faults, viz. single two line bridging faults in static CMOS combinational circuits is presented. This algorithm uses results from I _DDQ measurement based testing.Unlike known diagnosis algorithms, this algorithm does not use fault dictionaries, it uses only logic simulation and uses no fault simulation. It also uses SOPS, a novel representation of subsets of two-line bridging faults resulting in an efficient algorithm.In spite of the large number of faults that we consider, our experimental results point to the computational feasibility of I _DDQ Measurement based diagnosis of single two line bridging faults. It also shows the effectiveness of reducing the set of possible faults using I _DDQ measurements.A preliminary version of this work was presented at the 29th ACM/IEEE Design Automation Conference, 1992.Research Partially Supported by NSF Grant No. MIP-9102509.This work was performed when the author was with the Dept. of Computer Science, State University of New York at Buffalo.     

Biomaterials research and development

This paper discusses the importance of biomaterials technology and its wide ramifications in clinical practice. The current problems in biomaterials research such as characterisation, surface modification, safety evaluation and tissue-prosthetic bonding are out lined. The need for constant access to several discipline would seem to underline the importance of large integrated centres in the development of biomaterials technology.     

A tilting disc valve — component materials and hydraulic function

The component materials used in fabrication of the Chitra heart valve, their choice and screening are described. Further the haemodynamic performance of this valve, which is under development and an equal sized No. 27 Bjork-Shiley valve prosthesis was compared in a left-heart pulse duplicator under similar conditions of flow rates and pressures. They were tested in both the aortic and mitral positions of the duplicator. Regurgitant volumes and transvalvular pressure gradients were measured over flow rates ranging from 2 to 8 LPM. Flow patterns of the fluid flow across the valves were also photographed. The results indicate that the performance of the indigeneous valve is comparable, if not marginally better, to that of the well-established Bjork-Shiley valve. Transvalvular gradients and regurgitant volumes were marginally lower for the Chitra valve. This is attributed to the improved design of the valve disc shape.     

A characterization of robust test-pairs for stuck-open faults

Tests for stuck-open faults in static CMOS circuits consist of a sequence of two input vectors. Such test-pairs may be invalidated by delays in the circuit. Test-pairs that are not invalidated by delays in the circuit are known as robust test-pairs. We present a six-valued logic system Ω = {0, 1, r, f, 0h, 1h}. We show how Ω differs from a number of other logic systems that have been proposed for test generation. This logic system abstracts the important aspects of the transition behavior of the circuit, on application of an input pair, that is necessary to characterize robust test-pairs for stuck-open faults. This characterization of robust test-pairs is used to derive:

1. an algorithm for determining if a given test-pair is a robust test-pair for a given stuck-open fault or not; and
2. a simplified algorithm for computing a robust test-pair for a stuck-open fault. The resulting algorithm for computing robust tests for stuck-open faults can be implemented by minor modifications to test generation algorithms for stuck-at faults.

     

Training doctors for tomorrow

Enterotoxaemia of sheep and goats occurs worldwide, but the condition in goats is poorly understood. The disease in goats is mostly caused by Clostridium perfringens type D, although the role of the toxins of this microorganism in the pathogenesis of the disease is not fully understood. The disease occurs in three forms, peracute, acute and chronic, the cardinal clinical sign of the acute and chronic forms being diarrhoea. The main biochemical alterations are hyperglycaemia and glycosuria, while at necropsy the disease is often characterized by haemorrhagic colitis. The typical histological changes observed in the brain of sheep with enterotoxaemia are not considered to be a common feature of enterotoxaemia in goats. Although the pathogenesis of caprine enterotoxaemia has not yet been properly defined, it is usually accepted that the presence of C. perfringens type D in the small bowel, together with a sudden change to a diet rich in carbohydrates, is the main predisposing factor for the disease. Vaccination seems to be poorly effective in preventing caprine enterotoxaemia, which might be due to the fact that the enteric form of the disease is partially independent of circulating C. perfringens toxin. More studies are needed on caprine enterotoxaemia, especially of its pathogenesis and immunity, in order to develop more efficient control measures for this disease.     

Algorithms to select I DDQ measurement points to detect bridging faults

I _DDQ measurement is a time consuming process. Thus, reducing the number of I _DDQ measurements have a great impact on the test time. Carefully selecting a few I _DDQ measurement points is therefore an important problem. This problem has been studied for detecting leakage faults but not for bridging faults. We present novel algorithms to select I _DDQ measurement points to detect bridging faults. Experimental results obtained are very encouraging. The method can also be used: by test generators to compress I _DDQ test sets; and to maximize the fault coverage when a fixed number of measurement points are given.Research supported by NSF Grant No. MIP-9102509.