Toshihiro Taketa

An efficient information retrieval method in WWW using genetic algorithms

The information retrieval framework based on the self-organizing information retrieval network has the ability to search the information from any World-Wide Web (WWW) host. In order to improve the performance of this search method, we introduce the evolving information retrieval agents which are based on genetic algorithms (GAs). The evolving information retrieval agents which are generated in hosts automatically tune up their search parameters in order to adapt to the searching environment (machine specification, quality and quality of information, and so on) in the WWW host. The experiments show that the evolving information retrieval agents have suitable search parameters and make an efficient search from each starting WWW host.

Parallel FFT algorithms using radix 4 butterfly computation on an eight-neighbor processor array

Abstract Fast Fourier transform (FFT), which has wide and variety application areas, requires very high speed computation. Since parallel processing of FFT is very attractive for high speed FFT computation, many processor arrays and multiprocessor systems have been proposed with efficient FFT algorithms. As a result of the recent development of VLSI technology, several massively parallel computers have been implemented on commercial basis. The MasPar, which is one of the SIMD type massively parallel computers, consists of an eight-neighbor processor array. This paper discusses parallel 1-D FFT algorithms on an eight-neighbor processor array. We propose three algorithms according to various data allocation methods. Then we estimate and evaluate their processing time. With the number of processors N = N r × N r , processing time is estimated to be 2( N r − 2) t c + ( log 2 N r ) t b , where t c is the communication time between neighbor processors, and t b is the execution time for the radix 4 butterfly computation. We also compare these algorithms with the conventional radix 2 FFT algorithm implemented on a mesh processor array. It is shown that the radix 4 FFT algorithms are faster than the radix 2 algorithms. These algorithms get high speed FFT computation by combining the radix 4 FFT algorithm with the characteristics of the eight-neighbor processor array.

Multimedia and routing specific applications on IPv6 networks

The most promising applications of IPv6 networks might be to embed any network structure, logically, inside its vast addressing space. We show here some useful examples of such applications as a basic multimedia application of H.323 video conferencing, a method of phone number mapping, personal ID number mapping, machine ID number mapping all in the IPv6 address space, a method of multi-homing with IPv6 addressing, and that of IPv4 source address inclusion in IPv6 addresses. Some of the applications are implemented and evaluated.

Numerical backward simulation model with case branching capability

The authors are studying backward simulators which trace from results to causes for comprehensive verification of system safety. It is not easy to make a backward simulation model because the forward model may not be expressed in a reversible formula, or it is not reversible in cases of multiple inputs or inclusion of internal state variables. In this paper, we propose a backward simulator which can incorporate numerical simulation models and has a case branching capability to deal with multiple inputs. As a practical simulation target, we implemented a simulator for testing stability of dynamic pricing for power usage control as a smart grid application. We show some illustrating results of the backward simulation.

Object-oriented framework for cross-layer communication

The layered structure of network processing is an effective formalism for constructing software systems. However, it has weakness in the efficiency of the total network system. Recently, cross-layer communications among network objects are expected to be a savior, which patches the lack of functionality in the layered structure. This paper describes the necessary conditions which must be met for implementing the cross-layer communication, a proposed object-oriented framework for it, an abstract oriented representation of data, and its evaluation relating to the data efficiency. Also, some application examples are shown.

An idea of the agent-based information recommending system using the statistical information

The agent-based information recommending system we propose is based on statistical information and has the aim to provide the following information for users: the information the user is interested in; and the information the user potentially needs. The recommendation of the information the user is interested in is a service where the system recommends the selected information by users interest, concern and preference. The recommendation of the potential information the user needs is a service where the system watches the users daily browsing and searching, and deduces the information the user needs and recommends the new information the user seems to want. We describe an idea of the information recommending system and consider the design and the implementation.

Backward range simulation of Java bytecodes and reduction of its processing time

In this paper we introduce a backward simulation of machine code program and report results of reduction methods of its processing time. It may be efficient to analyze a program by tracing back from the final result for detecting all the bugs caused by some codes in the program. We describe the backward simulator of Java bytecode program and methods of processing time reduction such as range division, stepwise resolution improvement, setting of search path, and exponential range division. Results of two short programs and effects of reduction methods are shown.

Case Branching Backward Simulator for Integer Factorization

To verify the safety of a system, it is desirable to test all the situation, at least, by simulation. However, it is difficult to test the entire situation because the number of situations is almost uncountable. As a promising method, backward simulation which traces back from the undesirable result to causal input(s) is proposed. For such backward simulation, we have to transform the forward model into a backward model. This paper shows an example of transformation in the case of mathematical model of integer factorization, and shows the effectiveness of the backward simulation. Such backward simulation would require case branch processing. And the practicality of the simulation result will be influenced by the efficiency of the case division. We show an effective case branch processing by using stack and threaded actors which are promising schemes applicable to distributed processing.

Radix R parallel FFT algorithms with a global interconnection networks and its evaluation

In order to efficiently compute Fast Fourier transform (FFT) various parallel algorithms and their implementation to multiprocessors and multicomputers have been developed. In general, the local interconnection network is more high speed than a global one, but its capability depends on network architecture. On the other hand, the global interconnection network is not so high speed, but it does not depends on network architecture. It provides a flexible communication interface to the programmer. In this paper, we discuss parallel radix R FFT algorithms on a multiprocessor or multicomputer system with a global interconnection network. We propose two algorithms a stage-by-stage method and a multi-stage method. We also estimate the communication time. Then we show that the communication time is very sensitive to and affected by data exchange strategy. Finally, we implement these algorithms on two commercial massively parallel computers (nCUBE/2 and CM5) and measure the communication time.

Performance evaluation of high-speed self-token ring LAN

The fiber distributed data interface(FDDI) is now widely accepted as the follow-on LAN for IEEE 802.3 (the Ethernet) and 802.5 (the token ring) LANs. However, the advent of more high-speed LANs is eagerly expected to support higher performance requirements. In this paper, we describe a new ring access control scheme adopting multiple-tokens, referred to as the self-token protocol. In the protocol, each station has private tokens, called self-tokens, and has a fixed length register to prevent packets on a ring from collision. After approximate analysis of throughput-transfer delay characteristics, we show that this protocol is attractive and suitable for a gigabit LAN. We also show that fairness of this protocol is kept good for a low number of self-tokens.<<ETX>>