John C. Thompson

Formalization of operations and function definitions in a functional programming language for data structures

This paper presents a method for formalization of operations on certain class of syntactically represented linked Data Structures and then defines a set of functions to perform the operations. The formalization will be done by designing a set of transformation rules which are based on the grammars generating these Data Structures. Rooted in formal language theory, this method of formalization provides a basis for mathematical demonstration of the correctness of the operations. The function definitions are based on the idea of Functional Programming (FP) originated by John Backus. These function definitions will provide more functions for implementation of Extended Functional Programming (EFP) language designed for operations on Data Structures.

A demonstration of adaptive weather-surveillance capabilities on the national weather radar testbed phased-array radar

This paper describes the latest adaptive scanning capabilities of the National Weather Radar Testbed Phased-Array Radar located in Norman, OK. Focused observations, tailored observations, and the required scheduling algorithms are introduced, and their performance is illustrated with real-data examples. It is demonstrated that adaptive scanning for weather radars has the potential to reduce revisit times and to provide meteorological data that can aid in the forecasters warning-decision process.

Adaptive-Weather-Surveillance and Multifunction Capabilities of the National Weather Radar Testbed Phased Array Radar

The National Weather Radar Testbed (NWRT) is maintained and operated by NOAAs National Severe Storm Laboratory in Norman, OK, USA. It is a phased array radar (PAR) that was established to evaluate the potential to perform aircraft and weather surveillance with a single, multifunction radar. The NWRT PAR is also being used to demonstrate advanced weather-surveillance concepts that are well suited to the unique capabilities offered by phased arrays. This paper provides an overview of the adaptive-weather-surveillance and multifunction capabilities of this system.

G-base fourth normal form in relational database

G-base Fourth Normal Form is a Grammar based method used to formulate static and dynamic Fourth Normal Forms as a set of Wijngaarden grammars (W-Grammars) in relational database. Static W-grammar models static semantics (attribute level) which are considered to be time independent. Dynamic W-Grammars represent not only the interconnection in attribute level, but also consider the tuples of the relation and the relation of the components of the tuples to one another. The most important result is construction of parsers which are essentially of the class LR(1) for these grammars and thus verification can be done efficiently. Using such systems, a database administrator or a user could test the correctness of a proposed schema without the necessity of possessing a deep knowledge of database theory.

Implementation of extended functional programming language on cellular tree and data flow architecture

Extended functional programming language (EFP) is an extended version of FP[BJ78] which can simply manipulate the operations on syntactically represented data structures[SR87]. In EFP a complex operation such as insertion of node in a binary tree will be expressed in a set of parallel operations which are much simpler. Both CTA (cellular tree architecture) [MG80] and DFA (Data Flow Architecture) [RM83] can be used to implement the operations on data structures(in particular) in parallel. CTA which has a full binary tree structure contains an array of independent cells which are the leaves of the tree. Each cell will be involved with only one symbol of the syntactically represented data structure. Since these cells are independent, they can perform the operations on these symbols in parallel. So in CTA, we get parallelism within the operations as well as between the operations. For DFA, each complex operation will be converted to a sequence of primitive operations which may be performed in parallel. This sequence of operations will be represented by a basic block of instructions[RM83]. All the instructions with the number of dependencies equal to zero can be performed in parallel(if there are enough processors). Each instruction represents a primitive operation. Only one processor is involved with the execution of each instruction. So there is no parallelism within the operation but only between the operations. In [SR87] these two computer architectures have been briefly described and by giving some examples, it has been shown how a specific function can be implemented on these computer architectures. At the end, these two computer architectures have been compared and contrasted by discussing some advantages and disadvantages of each method.

Data flow relation processor for knowledge base machine

A PROLOG query to the Knowledge-Based System is evaluated by means of massive join operations. Three approaches of query compilation are discussed. Existing join modules are studied with respect to the time and hardware complexities. Data Flow Relation Processor (DFRP) is a high performance join module which is suitable for joining n consecutive relations, and i ts function and organization are proposed. In DFRP, only those tuples of two relations that agree on the attribute common to both relations are directly concatenated in parallel. As soon as a concatenated tuple is formed in a PE, i t is sent to the next relation. The intermediate results are not formed. The way that DFRP behaves to perform join operations over a large number of relations naturally corresponds to the parallel deductive search and unification of the theorem-proving. Deadend packets in DFRP do not affect the query response time, while the backtracking of the theorem-proving degrades the performance.

TRANSPORTABILITY OF DATA AND RELATED SOFTWARE IN ELECTRICAL DISTRIBUTION SYSTEM MODELING

This paper discusses some results of a research project whose goal was to develop better methodology for distribution system planning (Gonen and Thompson, 1979). One of the design objectives of the implemented system was to make its operation as independent as possible of the host hardware/software. In order to achieve this portability, given the present technology, it was necessary to simulate many of the functions of a computers operation system with the distribution system planning model (DSPM) itself (Thompson and Gonen, 1980). These functions are located in the shell program which provides an interactive interface between the distribution system planner and the functions and programs of the DSPM.

A NETWORK EDITOR AND SHELL PROGRAM FOR ELECTRICAL DISTRIBUTION SYSTEM PLANNING

Recent research has led to the development of new tools for the electric energy distribution system planner. Two of these new tools are discussed in the paper. The first is a Network Editor which allows the planner to manipulate network objects much as one manipulates text using a text editor. The second is the overseer program for the planning system which provides the planner many of the features of an operating system.