Search

This work presents a set of hands-on educational activities designed to teach some cyber security concepts in the classroom. The experimental configuration used an ad-hoc wireless and wired network, and a Raspberry Pi implementing a Web and an SSH server. Students were connected using their own devices (laptops or tablets). Initially the students tested DoS attacks to the Web server using various tools. Next, the students had to create SSH accounts to the server and a pair of RSA keys; using their SSH accounts, the students had to transfer their public keys to the server. Finally, students had to attack the SSH service from Kali Linux running on virtual machines in teams, each team using a different tool. The scenario was implemented in parts during a series of lessons and was positively accepted and evaluated by the students, who got familiar with a number of concepts and tools of computer networking and network security. In the end, the students informally assessed the Kali Linux SSH attack tools. Ways to assess the students qualitatively and quantitatively based on their participation are also presented. These lab exercises used a series of open source software, as well as low-cost equipment. [ABSTRACT FROM AUTHOR]

In the last decades the energy consumption of data centers has increased exponentially. The high energy consumption is due to power consumption of the IT resources, to the cooling system and to the inefficient usage of the resources. From the total power consumption in a data center one half is consumed by active servers and another half is consumed by the cooling system, network of the data center etc. There are several techniques for increasing the efficient usage of the resources in a data center. One of these techniques is the virtualization technology. This enables each user to work with his own virtual machine that has its own operating system. The purpose of the present paper is to present an optimization model that tries to find the best allocation of virtual machines to servers in a data center that minimizes the overall power consumption. The model has two objective functions. The first objective function is the IT power consumption of the servers. The second objective function is connected with the power consumption of the cooling system. The aim of the model is the minimization of the two objective functions. Constraints are given on the resource availability of the server resources (CPU, memory, bandwidth etc). The decision variables of the model are represented by a matrix with binary entries that describes the allocation of virtual machines to servers and by a vector with binary entries that indicates the active servers. In the paper is used a new idea for cooling power minimization. In order to do this the variance of the consumed power of the servers is minimized. This ensures the minimization of cooling power. Starting from the above multi-objective model we formulate several single objective models. [ABSTRACT FROM AUTHOR]

Dynamic resource allocation in cloud computing is the process of dynamically assigning resources to cloud applications according to current workloads. In the Infrastructure-as-a-service (IaaS) cloud computing service model, resources are allocated in the form of Virtual Machines (VMs) that can be resized and live migrated at runtime. The primary goal in a data center is to avoid wasting resources as a result of under- and over-utilization, which may result in violations of the Service Level Agreement (SLA) between the customers and the cloud provider. To satisfy the demands of users in a cost-effective way, the cloud providers must manage resources in an efficient manner. Our goal is to develop a distributed approach for dynamic resource allocation in cloud environments in order to maximize a global utility function. [ABSTRACT FROM AUTHOR]

The last decade has witnessed an exponential increasing in the number and the size of data centers. This expansion was driven by the increasing demand for digital products and services from the various sectors of the economy. Data center operators are faced with an increasing pressure for managing their facilities more efficiently and cost effectively such that they achieve 100 percent uptime. This is a difficult task to overcome, taking into account that energy consumption is the fastest growing component of data center. In an effort to answer this challenge, data center operators are using a virtualization technology. It enables each user to work with his own virtual machine that has its own operating system. Instead that the users have several independent computers, they run at the same time on the same server. Server consolidation is a technology that simplifies the administration of a data center. It improves energy efficiency by improving resource utilizations and reducing the number of active servers (physical machines) in the contemporary service-oriented data centers. We formulate several binary programming models for the static server consolidation problem that determine the best placement of virtual machines on servers (physical machines) in a data center. One model minimizes the energy consumption of the data center. In case that the servers are of the same type this is equivalent with the minimization of the number of active servers. Another model maximizes the use of server resources (CPU, memory, bandwidth etc). Decision variables are represented by a binary vector that shows which servers are active and by a binary matrix that shows the allocation of virtual machines to servers. [ABSTRACT FROM AUTHOR]

Virtual reality games and image processing Apps are examples of mobile cloud computing services (MCCS) common on Smartphones (SPs) nowadays, requiring intensive processing and/or wireless networking. The consequences are slow execution and huge battery consumption. Offloading the intensive computations of such Apps to a cloud based server can overcome such consequences. However, such offloading will introduce time delay and communication overheads. This paper proposes to do the offloading to nearby computing resources in a cooperative computation sharing network via short-range wireless connectivity. The proposed SCCOF reduces offloading response time and energy consumption overheads. SCCOF is supported by an intelligent cloud located controller that will form the cooperative resource sharing network on the go when needed, based on available devices in the vicinity, and will use the cloud if necessary. Upon the initiation of the MCCS service via the App, our controller will devise the offloaded VMs as well as the offloading network. A study test scenario was performed to evaluate the performance of SCCOF, resulting in saving of up to 16.2x in execution time and 57.25% energy. [ABSTRACT FROM AUTHOR]

Amid the rapid growth of Internet users, cybercrime is becoming one of the most challenging tasks for the systems and applications designers to deal with. The cybercrime threat is reflected in the increased number of cases and methods used by criminals. Systems based on cloud computing are natural targets due to their complexity (greater room for security weaknesses) and increasing popularity. Cloud computing is a modern technology that enables users to share resources in a virtual storage and computing environment. A cloud system is based on multiple physical servers. It provides a universal environment with a (large) number of Virtual Machines (VMs) that is available to many users accessing this system via the Internet. This form of access makes cloud systems weaker than physical networks. In order to prevent or minimize the number of attacks and in turn to secure data storage, any malicious behaviour such as external undesirable interventions should be rapidly identified and halted if possible. In this paper, we focus on the discovery of malicious behaviour via determining unwanted symptoms rather than via targeting particular malicious behaviours of the system directly. The main contribution of this paper consists in several new mechanisms for monitoring Virtual Machines and further experimental work targeting efficient ways of visiting VMs in order to discover malicious symptoms. We want to find the fastest and the best set of weights for visiting VMs. [ABSTRACT FROM AUTHOR]

Cloud computing is a popular computing paradigm which enables delivery of on-demand services over the internet. A cloud environment consists of multiple customers requesting for the available resources. Cloud vendors who offer Infrastructure as a Service should enable efficient management of the available resources. Proper scheduling in cloud enables the selection of best suitable resources for task execution. In the present scenario, allocating the resources efficiently is a challenging job. Service providers need to ensure that their resources are utilized properly. A new scheduling algorithm is proposed to address the challenge of task scheduling in cloud. The user tasks are prioritized. Based on the priority, the tasks are assigned to the Virtual Machines. The task with highest priority is assigned to a Virtual Machine with highest processing power. The proposed model is simulated using the CloudSim toolkit. The experimental results reveal that the proposed model results in lower execution time as the resources are allocated more efficiently. [ABSTRACT FROM PUBLISHER]

Development environments have put quite some effort into being open platforms, however rewriting our own tools as plug-ins we found that while host platform and other plug-ins are open for addition of feature they are typically closed to accessing data. We report from our own experience rewriting the Codemap tool as an Eclipse plug-in as well as from maintaining the Moose platform and argue why IDEs need to adopt an Open Data paradigm to become valuable targets for development of novel tools. We propose that data created by plug-ins need to default to being openly shared with other plug-ins through uniform data structures. Adding appropriate meta-information to this setup will allow other plug-ins to even make use of this data in unanticipated ways. We further argue that opening up data should not stop at the development platform but that programing languages and even virtual machines should also open up their internal data for the greater benefit of development tool and API designers. [ABSTRACT FROM PUBLISHER]

Flexibility is an important characteristic of cloud computing: users can dynamically lease resources only when needed. While this feature brings along various benefits, there are several drawbacks and among them performance unpredictability is a major issue. In this paper we investigate the effects of a new source of variability introduced by the flexible CPUs allocation policy supported by several IaaS providers. According to this policy, CPUs with different computing power may be allocated to the same VM type. We will show through experiments on Amazon EC2 platform that the range of CPUs, allocated transparently to the users, may lead to very different performance experienced by an application executed with the same VM type. To perform our studies we developed a benchmark application that can either run with a predetermined workload, or scale the service demand according to the computational power of the allocated CPU. We also executed some benchmarks of the well known DaCapo suite, that is composed by real world open source applications. The collected measurements illustrate that the fluctuations of performance due to the flexible allocation policy may be very high and unpredictable. [ABSTRACT FROM PUBLISHER]

Handheld mobile technology can help disaster relief workers and soldiers in the field with tasks such as speech and image recognition, natural language processing, decision-making, and mission planning. However, these applications are computation-intensive, take a heavy toll on battery power, and often rely on good connectivity to networks, limiting their practical usefulness in a crisis. This paper presents a reference architecture for mobile devices that overcomes these limitations by exploiting cloudlets -- VM-based code offload elements that are in single-hop proximity to mobile devices. [ABSTRACT FROM PUBLISHER]

This paper describes a successive and updated work of Revirt project which presents a virtual machine replay framework on Xen hyper visor. As both the commodity hardware and Xen hyper visor have been changed significantly since the first publication of Revirt, the initial implementation does not meet the needs of modern architecture any more. This paper presents an improved implementation of virtual machine execution replay system called CAS Motion. CAS Motion has three contributions. First, CAS Motion uses the performance monitor of Intel Core2 processor to construct time point of recorded events, which makes the event record more complete and precise. Second, CAS Motion can fully support multi-core hardware platform which is prevalent today. Third, CAS Motion is developed with more general architecture design, which makes it deployable on upstream Xen hyper visor and Dom0. Our experiments under a varity of workloads shows CAS Motion has low performance impact on monitored DomU. The growth of record log is also in acceptable range. [ABSTRACT FROM PUBLISHER]

This paper presents Virtual-Machines-MIB, a MIB (Management Information Base) directed to virtual machines management through SNMP (Simple Network Management Protocol). Virtual-Machines-MIB aims to define a standard interface of virtual machines management, allowing the management of several virtual machines monitors, like Xen, KVM and VMWare, with a common SNMP management tool. Different from previous virtual machines management MIBs, which allows the manager to perform only monitoring operations, Virtual-Machines-MIB allows to perform control operations, like create, delete, restart, turn on, pause and shut down virtual machines. It is also possible to use the proposed solution to change a virtual machine's name, amount of RAM, virtual CPU's and virtual storage drives. Practical results are presented using ordinary SNMP management tools performing KVM and Xen management. To do this, SNMP agents which support Virtual-Machines-MIB were developed and installed on KVM and Xen hosts. These SNMP agents are based on NET-SNMP public domain's agent, that was extended to support Virtual-Machines-MIB using libvirt API. [ABSTRACT FROM PUBLISHER]

Traditional Virtual Machines are over-provisioned to provide peak performance and waste a lot of system resources. In this paper, we propose and implement a placement strategy of Virtual Machines based on workload characteristic-s. In our approach, the virtual machines are placed into various groups after several iterations and matching based on the complementary of virtual machines' workloads. Requested resources are allocated to virtual machines placed in the same group according to the sum of individual resource requests. The experiment results show that the resource utilization of our approach was increased by 37.5% compared to traditional placement approaches, and it was increased by 12.5% compared with non-iterative matching approach. And we can conclude that our approach uses fewer physical machines and provides acceptable application performance. [ABSTRACT FROM PUBLISHER]

The next generation of scalable network simulators employ virtual machines (VMs) to act as high-fidelity models of traffic producer/consumer nodes in simulated networks. However, network simulations could be inaccurate if VMs are not scheduled according to virtual time, especially when many VMs are hosted per simulator core in a multi-core simulator environment. Since VMs are by default free-running, on the outset, it is not clear if, and to what extent, their untamed execution affects the results in simulated scenarios. Here, we provide the first quantitative basis for establishing the need for generalized virtual time scheduling of VMs in network simulators, based on an actual prototyped implementations. To exercise breadth, our system is tested with disparate applications: (a) a set of message passing parallel programs, (b) a computer worm propagation phenomenon, and (c) a mobile ad-hoc wireless network simulation. We define and use error metrics and benchmarks in scaled tests to empirically report the poor match of traditional, fairness-based VM scheduling to VM-based network simulation, and also clearly show the better performance of our simulation-specific scheduler, with up to 64 VMs hosted on a 12-core simulator node. [ABSTRACT FROM PUBLISHER]

Virtualisation is a prevalent technology in current computing. Among the many aspects of virtualisation, it can be employed to reduce hardware costs by server consolidation, implement "green computing" by reducing power consumption and as an underpinning process for cloud computing enabling the creation of a range of virtual networks and virtual supercomputers. This paper presents performance measurements for a cloning system known as iVIC that has been developed in Beihang University, China. In an extension to earlier work, it focuses on the factors the limit the number of clones that can be successfully started. IVIC creates clusters of virtual computers that can communicate with each other through virtual switch mechanisms. The virtual switches can also allow communication between the clone environment and the physical world. Testing has been undertaken to identify the limiting factors for creating and starting numbers of clone machines, measure the power consumption of the physical system and the computational performance capability of the clones. [ABSTRACT FROM PUBLISHER]

Service providers and enterprises capitalize on evolving cloud computing technologies when building datacenters for their own computing infrastructure and/or for providing services to others. Virtual machines and their migration are an important underlying technology of Infrastructure as a Service (IaaS) for building efficient cloud computing infrastructure on a cluster of servers. Autonomous migration of virtual machines is designed to increase the overall resource utilization on a cluster of servers. If a particular computing pattern caused imbalance that triggers migration, this pattern will be remembered or "learned" with its corresponding migration details for future use. In this paper, we use a proactive learning methodology that not only accumulates the past history of computing patterns and resulting migration decisions but more importantly searches all predefined possibilities for the most suitable decisions. We set up an experimental environment that consists of extensive real world benchmarking problems and a cluster of 16 physical machines each of which has on average eight virtual machines. We demonstrate through experimental results that our self regulated autonomous VM migration increases resource utilization of the servers on which cloud computing IaaS is currently running. [ABSTRACT FROM PUBLISHER]

Energy consumption has become major operational cost in data centers. Virtualization technology used in cloud computing platforms can improve energy efficiency and reduce costs. There are many ongoing research projects focusing on power management for virtualized cloud by making power-aware resource allocation and scheduling policies. However, there is a lack of VM power profiling method in such research, because the power consumption of an individual virtual machine (VM) cannot be measured directly by hardware power meter. In this paper, a novel power metering model is proposed for VMs in the cloud environment, based on online monitoring of system resource metrics, to estimate the power consumption of a physical server as well as one or more VMs running on it. By analyzing problems found in experiments, the model is improved to be the classified-piecewise ternary linear regression model which can achieve higher accuracy. In addition, the model is proved to be effective by running a variety of sample programs. The implementation of our model shows that it can achieve average estimation accuracy of more than 96% with low runtime overhead. [ABSTRACT FROM PUBLISHER]

Cloud computing has emerged in recent years as a major segment of the IT industry; however, security concerns remain the primary impediment to full-scale adoption. Leveraging properties of virtualization, virtual machine introspection (VMI) has yielded promising research for cloud security yet adoption of these approaches in production environments remains minimal due to a semantic gap: the extraction of high-level knowledge of the guest operating system's state from low-level artifacts collected out-of-VM. Within the field of forensic memory analysis (FMA), a similar semantic gap exists from the reconstruction of physical memory dumps. We implement a production oriented prototype utilizing designs that combines and narrows these semantic gaps in a modular framework to function as an intrusion detection system (IDS) detecting and defeating post-exploitation activity. [ABSTRACT FROM PUBLISHER]

For Infrastructure-as-a-Service (IaaS) clouds, stepping-stone attacks via hosted virtual machines (VMs) are critical. This type of attack uses compromised VMs as stepping stones for attacking the outside hosts. Not only compromised VMs but also IaaS providers are regarded as attackers. For self-protection, IaaS clouds should perform active response against stepping-stone attacks. However, it is difficult to stop only outgoing attacks at edge firewalls of clouds because edge firewalls can use only information in network packets. In this paper, we propose a new self-protection mechanism against stepping-stone attacks for IaaS clouds, which is called xFilter. xFilter is a packet filter running in the virtual machine monitor (VMM) underlying VMs and achieves pinpoint active response by using VM introspection. VM introspection enables xFilter in the VMM to obtain information on packet senders directly from the memory of VMs. When xFilter detects outgoing attacks, it automatically generates appropriate filtering rules with information on sender processes. Our experiments showed that xFilter could stop only outgoing attacks as much as possible. The performance degradation due to xFilter was less than 13% in usual cases. [ABSTRACT FROM PUBLISHER]

This paper describes extensions to the PAPI hardware counter library for virtual environments, called PAPI-V. The extensions support timing routines, I/O measurements, and processor counters. The PAPI-V extensions will allow application and tool developers to use a familiar interface to obtain relevant hardware performance monitoring information in virtual environments. [ABSTRACT FROM PUBLISHER]

Virtualisation is a prevalent technology in current computing. Virtualisation can reduce hardware costs by server consolidation, implement "green computing" by reducing power consumption and is an underpinning process for cloud computing enabling the creation of a range of virtual networks and virtual supercomputers. This paper presents performance measurements for a cloning system known as iVIC developed in Beihang University, China. iVIC creates clusters of virtual computers that can communicate with each other through virtual switch mechanisms. The virtual switches can also allow communication between the clone environment and the physical world. The paper presents the results of impartial testing to identify the limiting factors for creating and starting numbers of clone machines, measure the power consumption of the physical system and the computational performance capability of the clones. [ABSTRACT FROM PUBLISHER]

In this paper, we present an approach to provide different execution environments within Omnivore, our peer-to peer based scheduling system for operating a pool of unused desktop computers as a desktop Grid, and integrating the desktop Grid into a larger Grid environment via the Grid Way meta-scheduler. The proposed approach is based on extending Omnivore to support virtualization technologies. A new plug-in infrastructure allows us to plug in different kinds of execution modules (including virtual machines) and select the most adequate execution environment for a job at run time. Since each job runs in its own virtual machine containing the required operating system, software and data, administrative efforts are reduced, security is improved, ease of use of the resources is enhanced, and utilization of the resources is increased. Experimental results are presented to indicate that there is a small overhead in using the proposed approach, but for long running jobs it is negligible. [ABSTRACT FROM PUBLISHER]

Wireless Sensors have seen a lot of applications in our daily lives in the recent years. The market has been flooded with high end consumer electronics using wireless sensor technology. However, most of the current technologies require the sensor to be in the vicinity of the end-user application. There has been some study in the techniques for sensor provisioning and sharing for the large number of existing Wireless Sensor Networks. Virtualization of Wireless Sensor Networks (WSNs) is a step forward in exposing these WSNs to large user base from remote locations. However, there is still a huge gap in bringing together information available from heterogeneous, distributed resources of Wireless Sensor Networks to a non-localized user. In this work, we utilize IaaS paradigm of Cloud Computing in virtualization of sensor networks which gives the flexibility of handling heterogeneous systems. The system also enables a smart device user to access information generated by Wireless Sensors through the cloud via SaaS based design. This allows the system to take common computational tasks to be hosted as a service through the cloud. It frees the smart device user from running heavy applications for data processing and storing. Thus, system provides the smart device user a Cloud Enabled Wireless Sensor Network infrastructure. The system architecture provides the necessary features for it to be scalable and flexible, ensuring reliable sensor data transfer and processing through cloud infrastructure. We also present a small test bed implementation of the system. [ABSTRACT FROM PUBLISHER]

Cloud Computing has been steadily evolving during the past few years with new types of services, based on cloud platforms, being introduced regularly. It becomes apparent that, as the technology and the available tools progress, clouds will be tasked to support increasingly heavier workloads. Efficient ways of scheduling are therefore required in order to effectively utilize the available computing capacity. This capacity, of course, comes at a cost which must be balanced with performance requirements. Thus, the resource leasing system plays an equally important role as the job scheduler. This paper studies a scalable Cloud Computing system modeled around the Amazon EC2 architecture, with a workload model that offers fluctuating traffic characteristics. Different job dispatching heuristics were used in combination with gang-scheduling and job migrations. Also a complex resource manager was incorporated to maintain cost-efficiency. The system was studied through the use of a discrete event simulator in order to assess the effect of different workload configurations and resource leasing schemes on the overall performance and cost of the modeled system. The simulation results, present the advantages and disadvantages of the applied configurations and highlight that different or hybrid strategies are required to achieve the best balance between performance and cost. [ABSTRACT FROM PUBLISHER]

Today, virtualization technology is ubiquitously woven into nearly every technical field and conversation, taking place in the world of Information Technology (IT), because it can provide various benefits in terms of cost effectiveness, availability, hardware utilization, resource protection, remote access, and other capability enhancements. As a result, the implications of virtual computing environments become profound and drive a shift in the fundamentals of information systems design, operation, and management. However, virtualization also introduces new challenges and concerns related to implementing secure virtualized computing environments. Therefore, in this paper we first discuss common exploits of security properties in virtualized computing environments and analyze their security vulnerabilities from the perspective of attackers. Consequently, we identify and discuss the main areas of virtualized information system design and operation in which security concerns must be addressed. Finally, we present our recommendations and future trends for trusted virtualized computing environments. [ABSTRACT FROM PUBLISHER]

Network-booting is widely adopted in universities that have to maintain many client computers. In conventional network-booting systems, the primary bottleneck is the disk image distribution servers and the network to these servers. To eliminate this bottleneck, peer-to-peer (P2P) methods must work. However, existing P2P methods, including BitTorrent, do not work well for network-booting because they are highly optimized for distributing an entire large file, while network-booting requires certain parts of a large file. In the present study, aiming to solve the problems described above, we describe a new network-booting system that uses a P2P method. In our P2P-based network-booting system, a client node receives disk blocks not only from central distribution servers but also from the other client nodes that already have the demanded disk blocks. To the best of our knowledge, our network-booting system is the first effective implementation of a network-booting system that uses a P2P method in a local area network. Unlike conventional P2P systems, the proposed network-booting system can deal with demanded parts of a large disk image. We performed experiments with 112 client nodes in real classrooms on a university campus. The results of the experiments show that our implementation scales well as the number of client nodes grows. [ABSTRACT FROM PUBLISHER]

Field-programmable gate arrays (FPGAs) have recently been used as platforms to emulate SystemC descriptions. Emulation supports in-system testing using real input and output. We previously showed emulation speed to be competitive with SystemC simulations on a PC when the emulator uses acceleration engines. A limit on the number of acceleration engines that can fit on an emulation platform creates new online problems involving runtime decisions as to when to load a SystemC process into an acceleration engine. We define the online SystemC emulation acceleration problem. In contrast to previous works that focus on statically improving SystemC (and the more general event-driven) simulations, we utilize online heuristics to manage the use of a limited number of SystemC acceleration engines in an emulation framework, where the kernel must adapt and react to dynamically changing process and event queues. We test several online heuristics and show 9x improvement over microprocessor-only emulation and 5x over statically preloaded acceleration engines. We further improve emulation performance by 10-20% by adding kernel bypass connections between acceleration engines and by adapting the online heuristics to make use of those connections. [ABSTRACT FROM AUTHOR]

Web applications play a major role in various enterprise and cloud services. With the popularity of social networks and with the speed at which information can be disseminate around the globe, online systems need to face ever growing, unpredictable peak load events. Auto-scaling technique provides on-demand resources according to workload in cloud computing system. However, most of the existing solutions are subject to some of the following constraints: (1) replying on user-provided scaling metrics and threshold values, (2) employing the simple Majority Vote scaling algorithm, which is ineffective for scaling Web applications, and (3) lack of capability for predicting workload changes. In this work, we develop an auto-scaling system, WebScale, which is not subject to the aforementioned constraints, for managing resources for Web applications in data centers. We also compare the efficiency of different scaling algorithms for Web applications, and devise a new method for analyzing the trend of workload changes. The experiment results demonstrate that WebScale can keep the response time of Web applications low even when facing sudden load changing. [ABSTRACT FROM PUBLISHER]

The Nexos Project is a joint effort between Marquette University, the University of Buffalo, and the University of Mississippi to build curriculum materials and a supporting experimental laboratory for hands-on projects in computer systems courses. The approach focuses on inexpensive, flexible, commodity embedded hardware, freely available development and debugging tools, and a fresh implementation of a classic operating system, Embedded Xinu, that is ideal for student exploration. This paper describes an extension to the Nexos laboratory that includes a new target platform composed of Qemu virtual machines. Virtual Xinu addresses two challenges that limit the effectiveness of Nexos. First, potential faculty adopters have clearly indicated that even with the current minimal monetary cost of installation, the hardware modifications, and time investment remain troublesome factors that scare off interested educators. Second, overcoming the inherent complications that arise due to the shared subnet that result in students' projects interfering with each other in ways that are difficult to recreate, debug, and understand. Specifically, this paper discusses porting the Xinu operating systems to Qemu virtual hardware, developing the virtual networking platform, and results showing success using Virtual Xinu in the classroom during one semester of Operating Systems at the University of Mississippi. [ABSTRACT FROM PUBLISHER]

We use per-student virtual machines to allow new students to configure servers, thus enabling them to develop an understanding of the complex eStadium system. The outcomes include: student learning as the per-student virtual machines progress into software development and production machines supporting the eStadium game-day environment; the teamwork and leadership skills that evolve as students progress from initial learning to leadership roles in the creation of sophisticated applications; guidelines for instructors mentoring students through the process of building and maintaining a working production system; and, parallels with best-practice software and system development in industry. The use of peer-evaluations and social-network studies enable us to determine how the students interact with and learn from each other across years (sophomores through seniors). This cross year, cross experience-level learning process is essential for maintaining the technical and team continuity of the project. It also prepares students in a very realistic way for the software-development process in industry. [ABSTRACT FROM PUBLISHER]

An enhanced parallel toroidal lattice architecture (TLA) for simulations of large-scale neural networks is proposed. The TLA was implemented on a transputer array. A planar lattice architecture for neural network simulations (PLANNS) is also proposed as an improved version of TLA. The processor connections of this architecture are configured in a completely planar lattice structure, which is the most efficient structure for implementation using wafer-scale integration and increased parallel expandability. The performance of the PLANNS is almost proportional to the number of processors. Furthermore, this architecture exhibits great flexibility in simulating the various neural network architectures and a variety of neuron models [ABSTRACT FROM PUBLISHER]

A completely parallel, asynchronous, pulse-processing neural net has been described. This hardware simulator is fabricated with discrete electronic elements. A variety of basic biological parameters are efficiently incorporated in the hardware. It is noteworthy that the concept uses neural activity to modify synaptic weights and time delays. It follows that learning rules are embedded in the net topology [ABSTRACT FROM PUBLISHER]