Wireless sensor networks are the emerging information technology for monitoring equipment and ambient conditions of production lines in modern factories. A salient feature of wireless sensor networks is that nodes can relay sensing data to a factory data centre in a multihop fashion. Existing security solutions to sensor networks focus on hop-by-hop encryption and message authentication, as the wireless medium is of an open nature. From an end-to-end perspective, a challenging security problem arises when computer hackers compromise some nodes in a network and these adversarial nodes modify or drop data packets destined to a data centre. In this paper, we propose the Secure Quasimultipath Forwarding (SQF) scheme with the novel idea of dynamic en route retransmission upon encountering adversary nodes. Modelling techniques are employed to study the impact of the data-forwarding security design on packet delivery latency and energy consumption. The simulation results demonstrate that SQF circumvents adversaries with a several-fold reduction in sensor node energy consumption, compared to the existing multipath routing method.

Denial of Service (DoS)/Distributed DoS (DDoS) attack is an eminent threat to an Authentication Server (AS), which is used to guard access to firewalls, virtual private networks and resources connected by wired/wireless networks. In this paper, a new protocol called Identity-Based Privacy-Protected Access Control Filter (IPACF) is proposed to counter DoS/DDoS attacks. The IPACF is stateless for both user and AS since a user and responder must authenticate each other. The value and identity for authentication are changed in every frame. Thus, the privacy of both user and server is protected. The performance of the implementation is reported in this paper. In order to counter more DoS/DDoS attacks that issue fake requests, parallel processing technique is used to implement the AS. The performance comparison of dual server and single server is also reported. To study the capability of IPACF when facing massive DDoS attacks, simulations using OPNET for a network consisting of 1000 nodes with 10 Gbps pipe to the AS are carried out. The simulations show that the performance of AS has very little degradation in terms of packet latency and CPU utilisation for users. Queueing models are used to compare simulations and agreement between models and simulations is acceptable.

DoS/DDoS attack is a prevalent problem to an authentication server, which is used to guard firewalls, virtual private networks and wired/wireless networks. The sticking point of this problem is the computation needed to distinguish the legitimate users and attackers. In this paper, a new intrusion-resilient network security system is proposed to defend against DoS/DDoS attack. We implement this authentication scheme in Java and run the simulation using OPNET.

Linking factory floors to the Internet, coupled with the rapid deployment of wireless access networks, is initiating a new paradigm for factory automation-a corporate employee with a handheld computing device can have anytime, anywhere access to the latest factory floor information. Authentication between a factory database and a remote user is crucial for such paradigm; however, existing authentication protocols are inadequate to defend against strong adversaries with break-in capabilities. In this paper, we design and implement the Energy-Efficient and Intrusion-Resilient Authentication (ERA) protocol. Through a novel combination of hash chain,pin, and message authentication code (MAC), ERA can achieve the security self-recovery when strong adversaries compromise either a user's handheld device or a factory authentication server to obtain the authentication secrets. The technique of mutual MAC is proposed to defend against online pin-guessing attacks launched by strong adversaries. Furthermore, an optimization of tuning hash chain iteration is introduced to reduce energy consumption of a handheld device. Analytical and experimental results show that ERA provides a better security guarantee and incurs much less computation and communication overhead than the existing authentication protocols.

Replacing specialized industrial networks with the Internet is a growing trend in industrial informatics, where packets are used to transmit feedback and control signals between a plant and a controller. Today, denial of service (DoS) attacks cause significant disruptions to the Internet, which will threaten the operation of network-based control systems (NBCS). In this paper, we propose two queueing models to simulate the stochastic process of packet delay jitter and loss under DoS attacks. The motivation is to quantitatively investigate how these attacks degrade the performance of NBCS. The example control system consists of a proportional integral controller, a second-order plant, and two one-way delay vectors induced by attacks. The simulation results indicate that Model I attack (local network DoS attack) impairs the performance because a large number of NBCS packets are lost. Model II attack (nonlocal network DoS attack) deteriorates the performance or even destabilizes the system. In this case, the traffic for NBCS exhibits strong autocorrelation of delay jitter and packet loss. Mitigating measures based on packet filtering are discussed and shown to be capable of ameliorating the performance degradation.

Authentication is an essential premise for deploying wireless LAN inter-network roaming in the real world. The technical challenge lies in the fact that a visited network does not initially have the authentication credentials of a roaming user. Existing solutions require that authentication is performed via three-party communications: user to visited network to home network. In this paper, a new concept that mutual authentication between a visited network and a roaming user can be performed locally without the contact with user's home network is proposed. This new protocol not only satisfies the roaming authentication requirements, but also has efficiency in terms of authentication time delay.

A denial of service (DoS) attack is a well-known problem for networks. Although many research papers have addressed the problem of DoS attacks, some drawbacks still exit. The major problem is this: when a responder receives a request or key message, and system resources are required to verify that the request has been sent from a legitimate user, the additional computing time and memory storage involved provides an adversary the opportunity to launch a DoS attack. In order to prevent this attack, we present a new protocol called the identity-based dynamic access control filter (IDF), which checks every filter value on every received frame to determine if they have permission to enter the system and use system resources. In this paper we focus on wireless network protection, but the same concept can be applied to a wired network as well. A performance analysis of the critical step used in preventing a DoS attack is conducted and the result is compared with other public key-based schemes.

The Gamma interconnection network (GIN) is composed of 3*3 basic building blocks, with interconnecting patterns between stages following the plus-minus-2/sup i/ functions. The authors consider modifications to the GIN by altering the interconnecting patterns between stages so as to achieve high terminal reliability between any source-destination pair, resulting in the reliable GIN (REGIN). A type of REGIN's ensures totally disjoint paths in existence from any source to any destination, thereby capable of tolerating an arbitrary single fault. If several building blocks (i.e., 3*3 switches) are fabricated in one chip with very large scale integrated (VLSI) technology, the layout area and the pin count are less for the REGIN than for its GIN counterpart as a result of the change in the interconnecting patterns, giving rise to potential cost reduction. The terminal reliability of the REGIN is derived and compared with that of a compatible GIN. In addition, the performance of the REGIN is evaluated using simulation.>

Nonequilibrium self-consistent RF discharge simulations which use a Cray X-MP/24 supercomputer to increase the computation rate are described. The fluid transport equations are solved by a more accurate flux corrected transport (MAFCT) algorithm. A staggered-mesh method is used for the arrangement of distribution functions. Vectorization and multitasking methods are used to optimize the computation time required by the simulation. Consequently, the simulation code is fully vectorized and the speedup ratio between the original code and the fully vectorized code approaches 11.7. Moreover, the multitasking has a speedup ratio of 1.988 on the two-processor Cray X-MP/24. The simulation results are compared with other model results to demonstrate the advantage of MAFCT. These results are also compared with experimental results. A better understanding of RF discharge mechanisms is accomplished through the use of a Cray X-MP supercomputer, which provides a solution within a reasonable amount of turnaround time.>

A message-passing computer architecture called the modified mesh-connected parallel computer (MMCPC) is proposed and studied. The MMCPC is designed to be general-purpose parallel architecture suitable for wafer-scale integration. Generalized stochastic Petri nets (GSPNs) are used to model the behavior of the MMCPC. The GSPN performance modeling results show a need for a new processing element (PE). A PE architecture, able to handle data processing and message passing concurrently, is proposed, and the silicon overhead is estimated in comparison with transputerlike PEs. Based on the proposed PE, optimum sizes of the MMCPC for different program structures are derived. A two-dimensional fast Fourier transform problem is used as an example to demonstrate that the MMCPC is a cost-effective performance-enhancement architecture to a real problem.>

We simulate three neural networks on a vector multiprocrssor. The training time can be reduced significantly especially when the training data size is large. These three neural networks are: 1) the feedforward network, 2) the recurrent network and 3) the Hopfield network. The training algorithms are programmed in such a way to best utilize 1) the inherent parallelism in neural computing, and 2) the vector and concurrent operations available on the parallel machine. To prove the correctness of parallelized training algorithms, each neural network is trained to perform a specific function. The feedforward network is trained to perform the Fourier transform, the recurrent network is trained to predict the solution of a delay differential equation, the Hopfield network is trained to solve the traveling salesman problem. The machine we experiment with is the Alliant FX/80.

Parallel Monte Carlo techniques for simulating the evolution of an assembly of charged particles interacting with a background gas medium under the influence of the electrical field are presented. This simulation problem has inherent parallelism in nature. All the particles can be traced independently in a short time interval. We have overcome three major difficulties: 1) the number of particles to be simulated is increasing over time due to the ionization process; 2) the conditional branching statements do not inhibit multiprocessing through proper manual program tuning; 3) concurrency and vectorization are fully utilized through the new parallelized Monte Carlo method. The shared—memory vector multiprocessor Alliant FX/80 has been used for performance measurements. Significant speedup has been achieved.