- I am a retired scientist. My Main interest, next to Computer science ist to understand people and mankind. I want to ... moreI am a retired scientist. My Main interest, next to Computer science ist to understand people and mankind. I want to contribute to show people that the world will get better and mankind will increase its intelligence fast.edit
Abstract: One of the current trends in space missions is an increased autonomy of the spacecraft regardless whether it is a satellite, shuttle, space station or rover / robot. The causes are manifold but the main reason is the increased... more
Abstract: One of the current trends in space missions is an increased autonomy of the spacecraft regardless whether it is a satellite, shuttle, space station or rover / robot. The causes are manifold but the main reason is the increased amount of acquired data and the growing complexity ...
Research Interests:
DNA sequencing technologies keep getting faster and cheaper leading to massive availability of entire human genomes. This massive availability calls for better analysis tools with a potential to realize a shift from reactive to predictive... more
DNA sequencing technologies keep getting faster and cheaper leading to massive availability of entire human genomes. This massive availability calls for better analysis tools with a potential to realize a shift from reactive to predictive medicine. The challenge remains, since the entire human genomes need more space and processing power than that can be offered by a standard Desktop PC for their analysis. A background of key concepts surrounding the area of DNA analysis is given and a review of selected prominent algorithms used in this area. The significance of this paper would be to survey the concepts surrounding DNA analysis so as to provide a deep rooted understanding and knowledge transfer regarding existing approaches for DNA analysis using Burrows-Wheeler transform, Wavelet tree and their respective strengths and weaknesses. Consequent to this survey, the paper attempts to provide some directions for future research.
Research Interests:
As remote sensing technology progresses, spacecrafts are generating an increasing amount of sensor data over the years - much more than what the downlink data rate can handle. The need to reduce downlink data volume through onboard image... more
As remote sensing technology progresses, spacecrafts are generating an increasing amount of sensor data over the years - much more than what the downlink data rate can handle. The need to reduce downlink data volume through onboard image compression, ...
Research Interests:
ABSTRACT
Research Interests:
ABSTRACT We propose an algorithm to solve the Token Distribution problem, a static variant of the load balancing problem, on d-Dimensional, reconfigurable meshes with toroidal connections and side length n. No other algorithms have been... more
ABSTRACT We propose an algorithm to solve the Token Distribution problem, a static variant of the load balancing problem, on d-Dimensional, reconfigurable meshes with toroidal connections and side length n. No other algorithms have been proposed under this model of computation. We show that for token size T, the discrepancy Δ between the maximum and minimum number of tokens per PE can be reduced to 1 in at most In2nΔ(T +4d id) steps.
Research Interests:
We present in this paper an architectural design for a wavelet transform chip for use in real-time one-dimensional signal processing applications. Based on the observation that further levels of the wavelet transform require only as much... more
We present in this paper an architectural design for a wavelet transform chip for use in real-time one-dimensional signal processing applications. Based on the observation that further levels of the wavelet transform require only as much computation as the first level, our architecture requires only one row of processing elements to compute the complete transform. This is compared to previous
Research Interests:
ABSTRACT We study the wavelength problem and arc (edge) congestion problem for communicating permutation instances on a ring. We prove that the numbers of wavelengths in the directed case Gamma! w , in the undirected case w, and the arc... more
ABSTRACT We study the wavelength problem and arc (edge) congestion problem for communicating permutation instances on a ring. We prove that the numbers of wavelengths in the directed case Gamma! w , in the undirected case w, and the arc congestion Gamma! ß and the edge congestion ß for a permutation instance I 1 on an n-vertex ring C n satisfy: Gamma! w (C n ; I 1 ) ¸ n 3 ß ; Gamma! ß (C n ; I 1 ) ¸ n 4 ß ; w(C n ; I 1 ) ¸ n 2 ß ; ß(C n ; I 1 ) ¯ n 2 : All bounds are the best possible. Determining Gamma! w (C n ; I 1 ) and w(C n ; I 1 ) is NP-hard, which can be shown by a modification of the NP-hardness proof of the wavelength problem in rings, for general instances. 1 Introduction Optical networking is a very quickly developing new area of research. It is a key technology in communication networks and it is expected to dominate important applications such as video conferencing, scientific visualisation, real-time medical imaging, high speed supercomput...
Research Interests:
Research Interests:
Research Interests:
ABSTRACT Protein structure prediction is considered to be the holy grail of bioinformatics. Ab initio and homology modelling are two important groups of methods used in protein structure prediction. Amongst these, ab initio methods assume... more
ABSTRACT Protein structure prediction is considered to be the holy grail of bioinformatics. Ab initio and homology modelling are two important groups of methods used in protein structure prediction. Amongst these, ab initio methods assume that no previous knowledge about protein structures is required. On the other hand homology modelling is based on sequence similarity and uses information such as classification, structure, sequence and dihedral angles for prediction.
Research Interests:
Research Interests:
ABSTRACT We study the wavelength problem and arc (edge) congestion problem for communicating permutation instances on a ring. We prove that the numbers of wavelengths in the directed case Gamma! w , in the undirected case w, and the arc... more
ABSTRACT We study the wavelength problem and arc (edge) congestion problem for communicating permutation instances on a ring. We prove that the numbers of wavelengths in the directed case Gamma! w , in the undirected case w, and the arc congestion Gamma! ß and the edge congestion ß for a permutation instance I 1 on an n-vertex ring C n satisfy: Gamma! w (C n ; I 1 ) ¸ n 3 ß ; Gamma! ß (C n ; I 1 ) ¸ n 4 ß ; w(C n ; I 1 ) ¸ n 2 ß ; ß(C n ; I 1 ) ¯ n 2 : All bounds are the best possible. Determining Gamma! w (C n ; I 1 ) and w(C n ; I 1 ) is NP-hard, which can be shown by a modification of the NP-hardness proof of the wavelength problem in rings, for general instances. 1 Introduction Optical networking is a very quickly developing new area of research. It is a key technology in communication networks and it is expected to dominate important applications such as video conferencing, scientific visualisation, real-time medical imaging, high speed supercomput...
Research Interests:
... in order to find the optimum coloring for significantly larger graphs becomes too computationally expensive to permit performance analysis. ... 6. Sanjeev Arora, Carsten Lund, Rajeev Motwani,Madhu Sudan and Mario Szegedy: Proof... more
... in order to find the optimum coloring for significantly larger graphs becomes too computationally expensive to permit performance analysis. ... 6. Sanjeev Arora, Carsten Lund, Rajeev Motwani,Madhu Sudan and Mario Szegedy: Proof verification and intractability of approximation ...
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
ABSTRACT Presents an algorithm for Permutation Routing on a reconfigurable mesh based on the divide and conquer paradigm. The algorithm has a worst case running time of 4n+O(log2n) communication steps and requires a maximum queue size of... more
ABSTRACT Presents an algorithm for Permutation Routing on a reconfigurable mesh based on the divide and conquer paradigm. The algorithm has a worst case running time of 4n+O(log2n) communication steps and requires a maximum queue size of 5. The O(log2 n) term is associated with communication of single words which convey information about the distribution of packets. Hence, if the packet length is much larger than a single word, then the constant associated with this term is negligible
Research Interests:
A new parallel architecture is presented consisting of a toroidal memory and a processing element for each of its rows and columns. The standard Hough transform for line detection in digital images and an alternative approach to line... more
A new parallel architecture is presented consisting of a toroidal memory and a processing element for each of its rows and columns. The standard Hough transform for line detection in digital images and an alternative approach to line detection based on mathematical morphology are implemented on this architecture. In the alternative approach the M line angles to be detected are represented by M structuring elements, which are used to erode the image and hence reduce the average number of accumulator accesses per pixel from O(M) to O(1) for reasonable images. The choice of structuring elements allows this method to be tailored towards a more detailed image analysis than the standard Hough transform. While neither implementation is asymptotically optimal it is argued that for realistic images they outperform, by several orders of magnitude, any of the alternatives presented in the literature
Research Interests:
We study the wavelength problem and arc (edge) congestion problem for communicating permutation instances on a ring. We prove the best possible upper bounds on the number of wavelengths and arc (edge) congestion in both directed and... more
We study the wavelength problem and arc (edge) congestion problem for communicating permutation instances on a ring. We prove the best possible upper bounds on the number of wavelengths and arc (edge) congestion in both directed and undirected cases.
Research Interests:
Research Interests:
Research Interests:
The real-time manipulation and visualization of massive data-sets is an important capability in many applications ranging from the analysis of ftuid dynamics to flight simulation. The ability to visualise process mechanics or physicrd... more
The real-time manipulation and visualization of massive data-sets is an important capability in many applications ranging from the analysis of ftuid dynamics to flight simulation. The ability to visualise process mechanics or physicrd measurements facilitates analysis and
Research Interests:
ABSTRACT
Research Interests:
Research Interests:
Research Interests:
A solution to the token distribution problem is presented for the 2-dimensional reconfigurable mesh with restricted bus length. The algorithm is shown to be asymptotically worst-case optimal in reducing the discrepancy Delta between... more
A solution to the token distribution problem is presented for the 2-dimensional reconfigurable mesh with restricted bus length. The algorithm is shown to be asymptotically worst-case optimal in reducing the discrepancy Delta between maximum and minimum processor loads to ffi in optimal Theta((DeltaGammaffi) Delta n) time steps. The algorithm meets the time complexity of current state-of-the-art algorithms for sorting and permutation routing, but remains a factor of 2 from the bisection bound. Publishing Information Submitted: Australian Computer Science Conference, ACSC'97, Sydney Australia Author Information Heiko Schroder studied mathematics, physics, and computer science at CAU in Kiel (Germany) where he received his Ph.D. in computer science in 1977. Until 1989 he was Assistant Professor of Computer Science at CAU and Kansas University (U.S.A.) and Senior Research Fellow at ANU in Canberra (Australia). Until 1994 he was Professor of Microelectronics in Newcastle (Australia)...
We propose an algorithm to solve the Token Distribution problem, a static variant of the load balancing problem, on d-dimensional, reconfigurable meshes with toroidal connections and side length n. No other algorithms have been proposed... more
We propose an algorithm to solve the Token Distribution problem, a static variant of the load balancing problem, on d-dimensional, reconfigurable meshes with toroidal connections and side length n. No other algorithms have been proposed under this model of computation. We show that for token size T , the discrepancy Delta between the maximum and minimum number of tokens per PE can be reduced to 1 in at most 2nDelta(T + 4d) steps. 1 Introduction Token Distribution is a static variant of the load balancing problem. Given that each processor possesses some initial load, the goal of a token distribution algorithm is to distribute the tokens (tasks) so that the final load is close to the average load. A token distribution algorithm is judged by two parameters: the time T that the algorithm needs to produce the final (desired) distribution, and the maximal deviation ffi from the average load which it guarantees for all processors. One of the simplest and therefore one of the most practi...
Research Interests:
The token distribution problem is an important data distribution problem akin to the problems of sorting and routing, in which data elements, or tokens - must be evenly distributed amongst the processors of a parallel network. To date,... more
The token distribution problem is an important data distribution problem akin to the problems of sorting and routing, in which data elements, or tokens - must be evenly distributed amongst the processors of a parallel network. To date, all algorithms to solve this problem on the mesh and reconfigurable mesh architectures have lagged (in terms of the time required to obtain a solution) the state-of-the-art solutions to equivalent sorting and routing problems, despite there being evidence (a smaller bisection lower bound) that the problem of token distribution is inherently simpler than either of these. In this paper, we propose an algorithm whose performance is asymptotically equivalent to that of state-of-the-art sorting and permutation routing algorithms on reconfigurable meshes with non-serial busses, and performance asymptotically equivalent to the bisection bound for meshes with serial busses. Keywords : Token Distribution, Dimension Exchange, Ring, Linear Array 1 Introduction T...
This paper introduces the Complete Josephus Cube, a fault-tolerant class of the recently proposed Josephus Cube and proposes a cost-effective, fault-tolerant routing strategy for the Complete Josephus Cube. For a Complete Josephus Cube of... more
This paper introduces the Complete Josephus Cube, a fault-tolerant class of the recently proposed Josephus Cube and proposes a cost-effective, fault-tolerant routing strategy for the Complete Josephus Cube. For a Complete Josephus Cube of order r, the routing algorithm can tolerate up to (r+1) encountered component faults in its message path and generates routes that are both deadlock-free and livelock-free. The message is guaranteed to be optimally (respectively, sub-optimally) delivered within a maximum of r (respectively, 2r+1) hops. The message overhead incurred is only a single (r+2)-bit routing vector accompanying the message to be communicated
Research Interests:
ABSTRACT A frequently used method of representing words in the area of speech recognition is as a sequence of LPC (linear prediction coefficient) vectors, which are real-valued 12- to 16-dimensional vectors each representing a time slice... more
ABSTRACT A frequently used method of representing words in the area of speech recognition is as a sequence of LPC (linear prediction coefficient) vectors, which are real-valued 12- to 16-dimensional vectors each representing a time slice of the speech signal. It can be associated with particular phonems (or transitions between phonems). One method of transferring speech in a real-time environment is through the use of codebooks; i.e. a set of representatives of LPC vectors. The method proposed is based on the following idea: a sequence of N LPC vectors is produced out of a few minutes of speech. Then a distance matrix is calculated using any suitable distance function such as the Itakura distance. Then this distance matrix can be analyzed by repeatedly looking for rows with a maximal number of small-valued entries. These rows refer to clusters in the LPC vector space and thus can be assumed to be good representatives to be used as members of the codebook
Research Interests:
Hot-Potato routing is a very simple and powerful approach for processor communication in parallel networks and it is widely used in practice. For the permutation routing problem on meshes, which is the standard problem considered in the... more
Hot-Potato routing is a very simple and powerful approach for processor communication in parallel networks and it is widely used in practice. For the permutation routing problem on meshes, which is the standard problem considered in the literature, we develop new analysis techniques and double the efficiency of the best deterministic algorithms. We also present lower bounds for the greedy
Research Interests: Potato and Lower Bound
Research Interests:
ABSTRACT Instruction systolic arrays have been developed in order to combine the speed and simplicity of systolic arrays with the flexibility of MIMD parallel computer systems. Instruction systolic arrays are available as square arrays of... more
ABSTRACT Instruction systolic arrays have been developed in order to combine the speed and simplicity of systolic arrays with the flexibility of MIMD parallel computer systems. Instruction systolic arrays are available as square arrays of small RISC processors capable of performing integer and floating point arithmetic. In this paper we show, that the systolic control flow can be used for an efficient implementation of arithmetic operations on long operands, e.g. 1024 bits. The demand for long operand arithmetic arises in the field of cryptography. It is shown how the new arithmetic leads to a high-speed implementation for RSA encryption and decryption.
Research Interests:
Research Interests:
ABSTRACT Molecular biologists frequently compare an unknown protein sequence with a set of other known sequences (a database scan) to d etect functional similarities. Even though efficient dynamic programming algorithms exist for the... more
ABSTRACT Molecular biologists frequently compare an unknown protein sequence with a set of other known sequences (a database scan) to d etect functional similarities. Even though efficient dynamic programming algorithms exist for the problem, the required scanning time is still very high, and because of the e xponential database growth finding fast solutions is of highest importance to research in this area. In this paper we present a new approach to biosequence database scanning on the instruction systolic a rray to g ain h igh performance at low cost. To derive an efficient mapping onto this architecture, we designed a fine-grained parallel sequence comparison algorithm. This results in an implementation with significant runtime savings on Systola 1024, a parallel computer of this particular architecture.