Kamal S. Ali

Digital circuit design using FPGAs

Abstract Field Programmable Gate Arrays are generic devices that contain a vast number of basic digital components. The interconnections between these components are defined by the user. By specifying these interconnections different digital circuits may be realized. FPGAS are reprogrammable devices, making the change from one digital circuit to another as simple as downloading a new interconnection file. Furthermore the definition of the interconnection file may be carried out at high level machine languages, making the design and debugging of complex digital circuits rather simple. This publication discusses the use of FPGA as an alternate solution to ASICs. A review of the FPGA technology and its impact on the use of ASICs is carried out. This is followed by a projection of the impact of FPGA on digital technology. Finally methods on introducing FPGA technology to in the classroom and its impact on the quality of future engineers in considered. A case study of the FPGA impact on the Computer Engineering Technology Department of the University of Southern Mississippi will be considered.

The Internet as an educational tool

Abstract Searching for information in books or other printed forms could well be a thing of the past, as most information is readily available in electronic format in the Internet. Using a computer console, an Internet connection and the proper software a vast sea of information may be accessed with ease. The main requirements are but a computer and an Internet link. Teaching Internet access allows educators and students to remain on the cutting edge of technology. Using the Internet as a reference resource in teaching classes has the immediate advantage of allowing quick access to vast resources, also the Internet may be used as a communication tool. In both cases, as a resource or a communications tool, the Internet is superior to the conventional educational tools. However, using the Internet has some disadvantages, such as the limited control over plagiarism and the difficulty in verifying the reliability of information resources. When comparing the advantages and drawbacks of the Internet as an educational tool it is clear that its advantages by far outweigh its drawbacks.

Neural network approach to digital control

This paper starts with an overview of a classical PID controller design. An account of how Neural Networks may be incorporated to provide control is such a setup. The example used in this paper is the problem of controlling a High Frequency Acoustics Platform (HFAP) in-flight. The HFAP is towed by a ship and flown in the water behind the ship to acquire acoustic data reflected from the sea floor. The stability of such a platform is of prime importance to the accuracy of data collected. Using fight data from previous runs of the platform, a Neural Network is trained. The trained network is then used to predict the behavior of the platform. These predictions may then be directly translated to control signals minimizing the platforms spatial deviations. In this paper results form the trained Neural Network on predicting the behavior of the platform are displayed. Network prediction results illustrating the ability of the network to operate with partial input are displayed. Displaying these results in contrast with conventional controller results given the same input parameters emphasizes the importance of such a feature. Finally the use of different network architectures and the cost of using these network, in terms of computing power is investigated.

The undeterministic manipulation of solid models for robot program synthesis

Abstract In automatic robot program synthesis the number of variables that should be taken into consideration become prohibitively numerous. Due to the ambiguity and sheer size of items to be considered conventional computation methods cannot satisfactorily solve the problem. A Neural Network model that acquires data from a Solid Modeling data base, combines the completeness of information provided by solid modeling with the uncertainty encountered in the grouping process to perform geometrical classification of objects. The capabilities of Neural Networks to learn non-geometrical patterns in the grasping process, are yet to be achieved. Much progress needs to be made in both the neural model complexity and the computing machinery power before real intelligent program synthesis can be achieved.

Application specific integrated circuit design on a PC platform

Application Specific Integrated Circuits (ASICs) are most commonly designed on high end computer platforms, such as the SUN workstation. This paper describes a cost-effective method of designing ASICs on PCS. With the rapidly increasing power of the PC, and the ease of communication through the Internet, today it is possible to run some of popular VLSI CAD packages at a much lower cost. Today, at the University of Southern Mississippi the software package MAGIC has been used for that purpose on PC based stations. This paper starts with an overview of ASIC design on PC platforms and its impact on industry and academia. The software and hardware requirements and the techniques needed to acquire VLSI capabilities on a PC are examined. This is followed by a comparison of the available VLSI software in the market and its suitability for the class environment.

Self learning for autonomous systems

Abstract Learning is a key element in the strive for machine intelligence. Unsupervised learning is even more important for robots or autonomous systems that operate in remote environment away from human interactions, such as the case in the fully automated factory floor. To achieve unsupervised learning, a variety of models and techniques have been employed by investigators. In this paper some of the models, especially in the area of Neural Networks are compared and contrasted. Special consideration will be given self organizing maps (Kohonen Networks) [1,6]. A comparison of the Kohonen Networks and their biological counterparts is given. The introduction of these systems to increase the intelligence, and hence the autonomy of systems, is considered.

Refinement of algorithms for the real-time simulation of human movements in computer models

The purpose of this research is to develop natural movement algorithms that are more computationally efficient than those presently being used. This paper provides a new mathematical algorithm for solving the inverse kinematic problems needed to simulate the limb movements of a 3-D virtual human figure. This is accomplished by first decoupling the system of equations, then performing singular value decomposition (SVD) on the resulting matrix of constants coefficients.

Unsupervised learning for autonomous systems

This paper addresses the impact of neural networks on autonomous systems. Some neural network models are used to illustrate the effectiveness and suitability of these networks for space exploration. Fault tolerance and self learning capabilities of neural networks are used to illustrate such suitability. The advantages and disadvantages of the utilization of neural networks in autonomous systems are discussed and contrasted with the conventional systems currently in use.

Connectionist Approach For Robot Grasp Planning

In this paper we show that learning schemes can be utilized to generate robot grasping points. These learning schemes can be based on the geometrical similarity of objects and the functional similarity of tasks. This approach will drastically increase the speed of the search process and enrich the systems knowledge base. A Neural Network model that acquires data from a sold modeling data base is suggested. This model combines the completeness of information provided by solid modeling with the uncertainty encountered in the grouping process to perform geometrical classification of objects.

The role of geometric modeling in computer integrated manufacturing

Abstract This paper briefly describes the theories behind the different geometric modelling techniques. Through an examination of some of the important applications such as Numerical Control Machining, Computer Aided Design, Computer Aided Assembly, and Robot Programming and Simulation, we conclude the need for the integration of two or more modeling techniques in one system. This paper also addresses the problem of interfacing geometric modelers with other components of Computer Integrated Manufacturing. This discussion involves the problems of standardization. Finally some of the major challenges facing computer geometric modeling are defined as future areas for research such as the representation of both form and size tolerance in the model and the predicted interface with expert systems.