Human Activity Recognition (HAR) underpins applications in healthcare, rehabilitation, fitness tracking, and smart environments, yet existing deep learning approaches demand dataset-specific training, large labeled corpora, and significant computational resources. We introduce RAG-HAR, a training-free retrieval-augmented framework that leverages Large Language Models (LLMs) for HAR. RAG-HAR computes lightweight statistical descriptors, retrieves semantically similar samples from a vector database, and uses this contextual evidence to make LLM-based activity identification. We further enhance RAG-HAR by first applying prompt optimization and introducing an LLM-based activity descriptor that generates context-enriched vector databases for delivering accurate and highly relevant contextual information. Along with these mechanisms, RAG-HAR achieves state-of-the-art performance across six diverse HAR benchmarks. Most importantly, RAG-HAR attains these improvements without requiring model training or fine-tuning, emphasizing its robustness and practical applicability. RAG-HAR moves beyond known behaviors, enabling the recognition and meaningful labelling of multiple unseen human activities.

Graph embedding techniques remains the only alternative for Graph Neural Networks (GNNs) in machine learning for applications for which data in the form of graphs. GNNs, however, require substantial computational resources and are not able to provide certain advantages of conventional neural networks, such as transfer learning. The existing embedding based techniques, however, rely on random walk based strategies, which are not efficient in capturing the topology information, or use more structured strategies followed by feature engineering to achieve the required embedding. We propose VCNN (VirtualCoordinate based Neural Networks) technique that directly uses distances to an extremely small set (< 1 %) of random nodes (anchors) for network sampling and embedding, thus eliminating the feature engineering step. Experiments with two benchmark datasets show that a simple shallow neural network configuration using VC can achieve comparable performance to existing embedding-based techniques requiring deeper neural networks, thus also eliminating the need to experiment with multiple neural network configurations. This finding is particularly relevant for applications where computational efficiency is crucial. This research extends the machine learning toolbox for networks and data in the form of networks, highlighting the importance of strategic parameter use and model simplicity.

Graph-based data present unique challenges and opportunities for machine learning. Graph Neural Networks (GNNs), and especially those algorithms that capture graph topology through message passing for neighborhood aggregation, have been a leading solution. However, these networks often require substantial computational resources and may not optimally leverage the information contained in the graph’s topology, particularly for large-scale or complex graphs.We propose Topology Coordinate Neural Network (TCNN) and Directional Virtual Coordinate Neural Network (DVCNN) as novel and efficient alternatives to message passing GNNs, that directly leverage the graph’s topology, sidestepping the computational challenges presented by competing algorithms. Our proposed methods can be viewed as a reprise of classic techniques for graph embedding for neural network feature engineering, but they are novel in that our embedding techniques leverage ideas in Graph Coordinates (GC) that are lacking in current practice.Experimental results, benchmarked against the Open Graph Benchmark Leaderboard, demonstrate that TCNN and DVCNN achieve competitive or superior performance to message passing GNNs. For similar levels of accuracy and ROC-AUC, TCNN and DVCNN need far fewer trainable parameters than contenders of the OGBN Leaderboard. The proposed TCNN architecture requires fewer parameters than any neural network method currently listed in the OGBN Leaderboard for both OGBN-Proteins and OGBN-Products datasets. Conversely, our methods achieve higher performance for a similar number of trainable parameters. These results hold across diverse datasets and edge features, underscoring the robustness and generalizability of our methods. By providing an efficient and effective alternative to message passing GNNs, our work expands the toolbox of techniques for graph-based machine learning. A significantly lower number of tunable parameters for a given evaluation metric makes TCNN and DVCNN especially attractive for resource limited IoT/mobile devices and for reducing power consumption of ML models.

Immense growth of network usage and the associated proliferation of network, traffic, traffic classes, and diverse QoS requirements pose numerous challenges for network operators. Though data-driven approaches can provide better solutions for these challenges, limited data has been a barrier to developing those methods with high resiliency. In this work, we propose SyNIG (Synthetic Network Traffic Generation through Time Series Imaging), which utilizes Generative Adversarial Networks (GANs) for network traffic synthesis by converting time series data to a specific image format called GASF (Gramian Angular Summation Field). With GASF images we encode correlation between samples in 1D signals on a single 2D pixel map. Taking three types of network traffic; video streaming, accessing websites and IoT, we synthesize over 200,000 traces using over 40,000 original traces generalizing our method for different network traffic. We validate our method by demonstrating the fidelity of the synthetic data and applying them to several network related use cases showing improved performance.

Supervised machine learning is often used to detect phishing websites. However, the scarcity of phishing data for training purposes limits the classifier's performance. Further, machine learning algorithms are prone to adversarial attacks: small perturbations on attack data can bypass the classifier. These problems make machine learning less effective for phishing detection. We propose two Generative Adversarial Network (GAN) based approaches that synthesize phishing and legitimate samples to mimic real-world websites. Information about real-world datasets is obtained from ten publicly available phishing datasets which are used by the AAE (Adversarial Autoencoder) and WGAN (Wasserstein GAN) for generating synthetic data. Using both real and synthesized data, we demonstrate how to implement classifiers with higher performance and more resistance to adversarial attacks. We propose a set of hypotheses and validate them through experiments to demonstrate: (i) indistinguishability of synthesized samples from actual ones, (ii) susceptibility of classifiers to adversarial attacks, (iii) mitigating adversarial attacks by training on larger datasets that include correctly labeled synthesized samples, and (iv) better performance of classifiers trained on large datasets. Our AAE and WGAN have been trained on a wide range of datasets, making us optimistic about its widespread applicability.

Tracking individuals or groups based on their hidden and/or emergent behaviors is an indispensable task in homeland security, mental health evaluation, and consumer analytics. On-line and off-line communication patterns, behavior profiles and social relationships form complex dynamic evolving knowledge graphs. Investigative search involves capturing and mining such large-scale knowledge graphs for emergent profiles of interest. While graph databases facilitate efficient and scalable operations on complex heterogeneous graphs, dealing with incomplete, missing and/or inconsistent information and need for adaptive querying pose major challenges. We address these by proposing an inexact graph pattern matching method, which is implemented in a graph database with a scoring mechanism that helps identify hidden behavioral patterns. PINGS (Procedures forINvestigativeGraphSearch), a graph database library of procedures for investigative graph search is presented. Results presented demonstrate the capability of detecting individuals/groups meeting query criteria as well as the iterative query performance in graph databases. We evaluate our approach on three datasets: a synthetically generated radicalization dataset, a publicly available patient’s ICU hospitalization stays dataset, and a crime dataset. These varied datasets demonstrate the wide-range applicability and the enhanced effectiveness of observing suspicious or latent trends in investigative domains.

In this paper, we propose a novel method to make distance predictions in real-world social networks. As predicting missing distances is a difficult problem, we take a two-stage approach. Structural parameters for families of synthetic networks are first estimated from a small set of measurements of a real-world network and these synthetic networks are then used to pre-train the predictive neural networks. Since our model first searches for the most suitable synthetic graph parameters which can be used as an “oracle” to create arbitrarily large training data sets, we call our approach “Oracle Search Pre-training” (OSP). Such an approach enables us to evaluate the robustness of the autoencoder to artificial training data while making predictions on real-world social networks.For example, many real-world networks exhibit a Power law structure in their node degree distribution, so a Power law model can provide a foundation for the desired oracle to generate synthetic pre-training networks, if the appropriate Power law graph parameters can be estimated. Accordingly, we conduct experiments on real-world Facebook, Email, and Train Bombing networks and show that OSP outperforms models without pre-training, models pre-trained with inaccurate parameters, and other distance prediction schemes such as Low-rank Matrix Completion. In particular, we achieve a prediction error of less than one hop with only 1% of sampled distances from the social network. OSP can be easily extended to other domains such as random networks by choosing an appropriate model to generate synthetic training data, and therefore promises to impact many different network learning problems.

Unlike the traditional Internet application such as web browsing and peer-to-peer(P2P), video streaming has been dominating the global network traffic for the past few years, raising many challenges for network providers. With the popularity of interactive videos, a.k.a 360° videos, resource requirement for video streaming has been further increased. Prior identification of these video traffic is useful for effective provisioning of network resources, yet it is difficult due to the end-to-end encryption of data. However, with the recent advances in Machine Learning (ML) methods, prior identification of these resource-demanding traffic types has become viable. Nonetheless, they require more training data, without which leads to poor performance. Collecting more training data may also pose issues related to delayed training time. To remedy this problem, in this paper, we propose a novel Generative Adversarial Network (GAN) based data generation solution to synthesise video streaming data targeting 360°/normal video classification. Taking over 600 actual video traces and generating ≈ 30000 new traces, our post-classification results show that we can achieve 5 - 15% of accuracy improvement compared to only having actual traces.

Analysis of large-scale networks is hampered by limited data as complete network measurements are expensive or impossible to collect. We present an autoencoder based technique paired with pretraining, to predict missing topology information in ultra-sparsely sampled social networks. Randomly generated variations of Barabási-Albert and power law cluster graphs are used to pretrain a Hadamard Autoencoder. Pretrained neural network is then used to infer distances in social networks where only a very small fraction of intra-node distances are available. Model is evaluated on variations of Barabási-Albert and Powerlaw cluster graphs as well as on a real-world Facebook network. Results are compared with a deterministic Low-rank Matrix Completion (LMC) method as well as an autoencoder trained on partially observed data from the test-network. Results show that pretrained autoencoder far outperforms LMC when the number of distance samples available is less than 1%, while being competitive for higher fraction of samples.

Malicious actors often use phishing attacks to compromise legitimate users' credentials. Machine learning is a promising approach for phishing detection. While the accuracy of machine learning algorithms is often dependent on the training data, very little attack data for training is available. We propose an approach for augmenting existing datasets that can be used by machine learning algorithms. We use an Adversarial Autoencoder (AAE) to generate samples that mimic the phishing websites and provide metrics to assess the quality of the generated samples. We test these samples against models trained with real-world data. Some of generated samples are able to evade existing detection model. We then use a portion of these samples in training. The new machine learning models are more robust and have higher accuracy. In other words, real-world phishing site data augmented with AAE synthesized data used for training the model is more effective for phishing detection.

Extracting connectivity information in massive social networks is important for many applications. Algorithms developed for undirected networks cannot be used with social networks characterized by directed edges. We present a method to extract the network topology from a small sample of distance measures without the need for exhaustive measurements. Tolerating missing data is also necessary when certain nodes participate in message passing, but are not accessible for direct measurements. An anchor-based sampling approach is proposed and compared with random sampling. As we demonstrate, real-world directed social networks such as Twitter and US election blogs have hop-distance matrices that are low-rank. Low-rank matrix completion techniques are thus used to recover the complete topology from a relatively small set of measurements. Evaluation of the proposed technique using metrics such as distance distribution, degree distribution, and hop-distances show that the proposed technique is effective even when only a small fraction of distance entries are available.

The search in graph databases for individuals or entities undertaking latent or emergent behaviors has applicability in the areas of homeland security, consumer analytics, behavioral health, and cybersecurity. In this setting, even partial matches to hypothesized indicators are worthy of further investigation, and analysts in these domains aim to identify and maintain awareness of entities that either fully or partially match the queried attributes over time. We provide a comprehensive version of a graph pattern matching technique called Investigative Search for Graph Trajectories (INSiGHT) to find emergent patterns of behaviors in networks and tailor the application to detecting radicalization in the homeland security domain. To enable analysts' accounting of recurring behavioral indicators and the recency of behaviors as the imminence of a threat, we provide parameterized methods to score multiple occurrences of indicators and to dampen the significance of indicators over time, respectively. Additionally, we provide an indicator categorization scheme and a match filtering technique to ensure that partial matches to the most salient indicators are identified while reducing the number of false positives. Furthermore, since individuals may be radicalized in small groups or be involved in collective terrorist plots, we introduce a non-combinatorial neighborhood matching technique that enables analysts to use INSiGHT to identify potential query matches from clusters of individuals who may be operating in conspiracies. We demonstrate the performance of our approach using a synthetic radicalization data set and a large, real-world data set of the BlogCatalog social network.

For many important network types, physical coordinate systems and physical distances are either difficult to discern or inapplicable. Accordingly, coordinate systems and characterizations based on hop-distance measurements, such as Topology Preserving Maps (TPMs) and Virtual-Coordinate (VC) systems are attractive alternatives to geographic coordinates for many network algorithms. We present an approach to recover geometric and topological properties of a network with a small set of distance measurements. The approach is a combination of shortest path (often called geodesic) recovery concepts and low-rank matrix completion, generalized to the case of hop-distances in graphs. Results for sensor networks embedded in 2-D and 3-D spaces as well as for social networks indicate that the method can accurately capture the network connectivity with a small set of measurements. TPM generation can now be based on various context appropriate measurements or VC systems instead of distances to a set of global anchors. The proposed method is a significant generalization that allows the topology to be extracted from a random set of graph shortest paths, making it applicable in contexts such as social networks where VC generation may not be possible.

A novel structured P2P architecture is presented for multi-attribute query search. The architecture uses a Ring-of-Rings topology, combined with a Bloom-filter based resolution scheme and a caching strategy to efficiently resolve queries for correlated attributes. The overlapped ring mechanism allows the system to adapt to the correlations among contents. Results indicate that using caching scheme and overlapped ring architecture produces nearly 63% reduction in latency in response, 67% reduction in hop counts and 16.4% reduction in cost of communication compared to the conventional methodologies. The results obtained in our proposed methodology has logarithmic time complexity and constant space complexity.

The use of directional antennas in mobile wireless networks has emerged recently due to their numerous advantages such as high gain, low interference and low transmission power to reach the same distance as omni antennas. However, directional antennas limit the coverage and connectivity in the network. In this paper, we propose an algorithm to calculate a topology preserving map for wireless sensor networks using directional antennas. In the literature, topology mapping have been proposed for sensor networks equipped with omni antennas; however, to the best of our knowledge none of the existing work has considered the use of directional antennas in sensor network topology mapping. The application of existing algorithms in directional sensor networks will consume significantly high energy, as it requires transmitting the number of sectors times more beacon messages to cover the omini antenna range. Thus in this paper, a novel beacon forwarding algorithm limiting the number of forwarding sectors based on packet received direction is proposed. The simulation results show that it is possible to generate an accurate topology preserving map can be generated by only forwarding beacons in total of 135° angle. Moreover, the energy and the number of edges used to generate the topology map can be reduced by nearly 50% compared to the topology map generation with omni antennas based on received signal strength and hop count.

Millimeter wave communication shows promise in realizing next generation wireless sensor networks for bandwidth demanding applications. Despite its support of multi Gbps data rates, MmWaves requires unobstructed line-of-sight and suffers from heavy path losses. Overcoming these in complex 3D environments requires sectored antenna arrays with narrow beam widths and adaptive beamforming. Therefore, network topology maps would be significant than ever in millimeter wave sensor networks. Traditional topology mapping algorithms rely on omnidirectional transmission and reception and are thus not tailored to such networks. A novel topology mapping algorithm, Millimeter Wave Topology Map (MmTM) is proposed for 3D deployments, which take advantage of the directional information available from beamforming antennas as well as their beam steering capability. An autonomous robot traverses the network recording the packet reception from different nodes along with the receiving antenna sector ID that delivers the packet with highest signal quality. The techniques used in standard IEEE 802.11ad protocol are used for the optimum sector selection and collision avoidance. MmTM is evaluated using two realistic sensor network environments and compared with prominent localization approaches based on received signal strength and hop count. The results show that proposed algorithm has a less than 0.7m distance error and more than 50% of nodes are located in the correct direction, which is 7m and 35% improvement in distance error and sector displacement matrices compared to other algorithms.

This paper proposes the use of graph pattern matching for investigative graph search, which is the process of searching for and prioritizing persons of interest who may exhibit part or all of a pattern of suspicious behaviors or connections. While there are a variety of applications, our principal motivation is to aid law enforcement in the detection of homegrown violent extremists. We introduce investigative simulation, which consists of several necessary extensions to the existing dual simulation graph pattern matching scheme in order to make it appropriate for intelligence analysts and law enforcement officials. Specifically, we impose a categorical label structure on nodes consistent with the nature of indicators in investigations, as well as prune or complete search results to ensure sensibility and usefulness of partial matches to analysts. Lastly, we introduce a natural top-k ranking scheme that can help analysts prioritize investigative efforts. We demonstrate performance of investigative simulation on a real-world large dataset.

Investigative graph search is the process of searching for and prioritizing entities of interest that may exhibit part or all of a pattern of attributes or connections for a latent behavior. In this work we formulate a related sub-problem of determining the pattern matching trajectories of such entities. The goal is to not only provide analysts with the ability to find full or partial matches against a query pattern, but also a means to quantify the pace of the appearance of the indicators. This technology has a variety of potential applications such as aiding in the detection of homegrown violent extremists before they carry out acts of domestic terrorism, detecting signs for post-traumatic stress in veterans, or tracking potential customer activities and experiences along a consumer journey. We propose a vectorized graph pattern matching approach that calculates the multi-hop class similarities between nodes in query and data graphs over time. By tracking partial match trajectories, we provide another dimension of analysis in investigative graph searches to highlight entities on a pathway towards a pattern of a latent behavior. We demonstrate the performance of our approach on a real-world BlogCatalog dataset of over 470K nodes and 4 million edges, where 98.56% of nodes and 99.65% of edges were filtered out with preprocessing steps, and successfully detected the trajectory of the top 1,327 nodes towards a query pattern.

While physical coordinates are useful for IoT and sensor network operations, physical localization is not a viable option for large-scale networks of simple devices in complex or harsh environments. Topology Preserving Maps (TPM) extracted from anchor-based Virtual Coordinates (VCs) are an attractive localization free alternative for physical maps. We present an approach, based on the theory of low-rank matrix completion, to extract TPMs with only partial information about VCs. Evaluation using 2D and 3D networks with random anchors shows that accurate TPMs can be obtained even when up to 40% to 60% of random coordinates are missing. Coordinate generation and communication cost thus may be reduced significantly. TPM generation can now also be based on a different VC system as long as it characterizes each node with distances to a small set of random nodes instead of a global set of anchors.

This paper outlines our on-going efforts to address the radicalization detection problem, the automated or semi-automated task of dynamically detecting and tracking behavioral changes in individuals who undergo the process of increasingly espousing jihadist beliefs and transition to the use of violent action in support of those beliefs. Leveraging the notion that personal trajectories towards violent radicalization exist, we take a graph pattern matching approach to track individual-level indicators using data fused from available public and government/law enforcement databases. We show that our approach provides analysts with the ability to find full or partial matches against a query pattern of radicalization, and a means to quantify the pace of the appearance of the indicators that may help prioritize investigative efforts and resources to prevent planned attacks.

A Topology Preserving Map (TPM) is an easily obtainable localization free connectivity based map that preserves physical layout features of 2D/3D sensor networks. This paper considers how to obtain physical maps and physical coordinates from a TPM when the physical locations of a subset of nodes are known. First, we present a General Procrustes Alignment (GPA) based solution, which is the optimal linear transformation solution to the problem. Second approach is based on thin-plate spline (TPS), which transforms the set of topology coordinates to physical coordinates using radial basis functions. Five representative 2D network topologies are used to evaluate and compare the TPS approach with GPA approach, and also with the existing distance vector-hop (DV-Hop) technique. Results are presented for the cases where the reference nodes are selected randomly from the entire network, or randomly from the inner and outer boundaries of the network. The results show that with less than 10% of nodes as reference nodes, a map with an average error less than 0.6 of radio range can be achieved with the TPS approach, which significantly outperforms both the GPA and the DV-Hop. TPS approach generalizes directly to 3-D networks as well.

Topology maps represent the layout arrangement of nodes while maintaining the connectivity. As it is extracted using connectivity information only, it does not accurately represent the physical layout such as physical voids, shape, and relative distances among physical positions of sensor nodes. A novel concept Maximum Likelihood-Topology Maps for Wireless Sensor Networks is presented. As it is based on a packet reception probability function, which is sensitive to the distance, it represents the physical layout more accurately. In this paper, we use a binary matrix recorded by a mobile robot representing the reception of packets from sensor nodes by the mobile robot at different locations along the robots trajectory. Maximum likelihood topology coordinates are then extracted from the binary matrix by using a packet receiving probability function. Also, the robot trajectory is automated to avoid the obstacles and cover the entire network within least possible amount of time. The result shows that our algorithm generates topology maps for various network shapes under different environmental conditions accurately, and that it outperforms the existing algorithms by representing the physical layout of the network more accurately.

In a wireless sensor network, virtual coordinates provide most of the advantages of geographic routing strategies without actually relying on the location information of the nodes. Using a mobile sink provides advantages such as distributing energy consumption throughout the network. However, nodes need to be updated about the new virtual coordinate of the mobile sink as it moves. In this paper, we propose Circular Update-Directional Virtual Coordinate Routing (CU-DVCR), an algorithm specialized in routing towards a mobile sink in virtual coordinates. Through a set of experimental studies we show that CU-DVCR consumes less energy compared to alternative algorithms while providing comparable performance.

Virtual Coordinate (VC) based Wireless Sensor Networks (WSNs) are susceptible to attacks resulting from malicious modification of VCs of individual nodes. While the impact of some such attacks is localized, others such as coordinate deflation and wormholes (tunneling) can cause severe disruptions. A comprehensive solution for detection of such attacks is presented that combines Beta Reputation System and a reputation based routing scheme. Latter ensures safe communication that bypasses malicious nodes during detection process. Attacks are identified by detecting changes in shape of network using topology maps. The topology distortion is rated using clusters identifiable by existing VCs thus requiring low computation and communication overhead. Simulation based evaluations demonstrate that the scheme efficiently detects deflation and wormhole attacks. The detection scheme easily differentiates between the changes in the network due to node failures, e.g., caused by battery drain, from those due to an attack.

Collaborative peer-to-peer (P2P), grid, and cloud computing rely on resource discovery (RD) solutions to aggregate groups of multi-attribute, dynamic, and distributed resources. However, specific characteristics of real-world resources and queries, and their impact on P2P-based RD, are largely unknown. We analyze the characteristics of resources and queries using data from four real-world systems. These characteristics are then used to qualitatively and quantitatively evaluate the fundamental design choices for P2P-based multi-attribute RD. The datasets exhibit several noteworthy features that affect the performance. For example, compared to uniform queries, real-world queries are relatively easier to resolve using unstructured, superpeer, and single-attribute-dominated query-based structured P2P solutions, as queries mostly specify only a small subset of the available attributes and large ranges of attribute values. However, all the solutions are prone to significant load balancing issues, as the resources and queries are highly skewed and correlated. The implications of our findings for improving RD solutions are also discussed.

A scalable network fault localization scheme based on compressive sensing is proposed. Aimed at large networks, the proposed scheme monitors a network with a few paths covering the network, and upon detection of anomalies in one or more paths, adaptively carries out additional end-to-end measurements to localize the faulty links. Each adaptive measurement covers a set of links identified based on the previous resolution. The scheme is highly scalable as the total number of measurements required grows logarithmically with the number of links in the network - a level of scalability not practically achieved for network data inference with compressive sensing so far. The scheme is tested on realistic Internet topologies with Gilbert-Elliott loss model calibrated with measurements made on Planet-Lab infrastructure. Results indicate that the converged solution of the proposed scheme achieves over 99% detection rates and less than 1% false positive rates. The proposed scalable scheme is accurate in terms of detection, cost effective in terms of implementation, and casts a minimal monitoring traffic load.

Geographical routing can provide significant advantages in wireless sensor networks. However in many sensor networks, it is difficult or costly to find the exact location of the nodes. The virtual coordinate techniques allow a network to acquire a coordinate system without relying on geographical location. In this paper, we describe MS-DVCR, an extension of a state-of-the-art virtual coordinate routing protocol (DVCR) with the ability to route towards a mobile sink. We describe the design principles and implementation of the proposed protocol and through an experimental study, we show that it matches the performance of a simple extension of DVCR for mobile sinks while providing a significantly lower energy consumption.

Inherent differences in geometric, deployment and propagation constraints prevent many 2D sensor network routing algorithms from scaling to 3D. Complex 3D shapes and surfaces make geometric coordinate based techniques inefficient or impractical for 3D. Geo-Logical Routing (GLR) switches between virtual and topology coordinate domains to overcome the local minima of each other. GLR is demonstrated to be highly effective for complex 3D networks. Addition of a second initial anchor pair in Extreme Node Search (ENS) increases the number of anchors, thus enhancing performance in networks with very high node degree, while traditional 2D ENS performs just as well in other cases. Simulations demonstrate that GLR with ENS anchor placement significantly outperforms the Greedy Distributed Spanning Tree Routing (3D-GDSTR) algorithm, but without the need for localization.

A method for obtaining topology-preserving maps (TPMs) from virtual coordinates (VCs) of wireless sensor networks is presented. In a virtual coordinate system (VCS), a node is identified by a vector containing its distances, in hops, to a small subset of nodes called anchors. Layout information such as physical voids, shape, and even relative physical positions of sensor nodes with respect to x- y directions are absent in a VCS description. The proposed technique uses Singular Value Decomposition to isolate dominant radial information and to extract topological information from the VCS for networks deployed on 2-D/3-D surfaces and in 3-D volumes. The transformation required for TPM extraction can be generated using the coordinates of a subset of nodes, resulting in sensor-network-friendly implementation alternatives. TPMs of networks representing a variety of topologies are extracted. Topology preservation error ( ETP), a metric that accounts for both the number and degree of node flips, is defined and used to evaluate 2-D TPMs. The techniques extract TPMs with ( ETP) less than 2%. Topology coordinates provide an economical alternative to physical coordinates for many sensor networking algorithms.

A Compressive Sensing (CS) based solution is proposed for centralized and distributed discovery of physical phenomena in large scale Wireless Sensor Networks (WSNs). WSNs monitoring environmental phenomena over large geographic areas collect measurements from a large number of distributed sensors. Compressive Sensing provides an effective means of discovery and reconstruction of functions with only a subset of samples. Traditional CS relies on uniformly distributed samples which limits practicality of CS based recovery. To enhance the flexibility of sampling and implementation, the proposed approach uses random walk based samples. Unlike uniform sampling, random walk based sampling enables individual nodes achieve phenomenon awareness, i.e., the physical distribution of the phenomenon. We also derive a theoretical upper bound for the reconstruction failure probability. Simulation results on the number of samples required and error show that random walk based sampling is comparable to uniform sampling but with superior energy efficiency. More importantly, the proposed scheme provides a practical solution for a range of applications where uniform sampling is less economical or even infeasible.

Existing methods for detection and tracking of mobile nodes and objects via sensor networks depend on physical localization of sensor nodes. Consequently, they suffer from disadvantages associated with localization, which is often costly, error prone, and in some cases even infeasible. Tracking mobile nodes in a topological coordinate domain is proposed as an alternative to that in geographic coordinate domain. Topological coordinates (TCs), the basis of recent developments in Topology Preserving Maps (TPMs), are derived from hop distances from each node to a small subset of nodes. A technique is presented to reduce the distortion closer to the edges of principal component based TPMs, thereby enhancing its accuracy, which in turn facilitates accurate mobility tracking. Topological Coordinate based Tracking and Prediction (TCTP) algorithm is proposed for tracking and predicting the position of mobile nodes in the TC domain. Simulation results presented show that tracking and detecting performance of TCTP is competitive with geographic coordinate based algorithms despite the fact that it relies only on hop distances to a subset of nodes.

Welcome to the IEEE Conference on Local Computer Networks, the conference on the leading edge of computer networking. Now in its 38th year, IEEE LCN is one of the oldest and continuously running conferences on computer networking. Back in Sydney after 8 years, we will continue with the stimulating and friendly interactive environment that LCN is known for, fostering interactions among researchers and practitioners, from academia and industry. LCN will continue its tradition of offering a quality set of papers selected via a thorough review process with a very selective acceptance rate. Poster and demonstration sessions will provide a window into leading edge research and development activities around the globe. Nine workshops will be held on the first and last days of the conference. LCN is a truly international conference with authors coming from 40 countries. LCN 2013 will offer keynotes from two distinguished leaders in networking: Kevin Bloch, CTO of Cisco Australia and New Zealand, and Prof. Aruna Seneviratne of University of New South Wales and National ICT Australia (NICTA).

Packet reordering is a common, and an unavoidable network phenomenon. Recovery from reordering requires re-sequencing buffers, e.g., at transport or application layer at the end nodes. Input or output buffers are used in high-performance routers to mitigate reordering due to massive internal parallelism. Reordering caused by multi-path routing in sensor networks has to be dealt with at nodes with limited buffers. Yet, no formal techniques exist to analyze or predict the improvement in packet order achievable with limited re-sequencing buffers. A theoretical foundation for analysis of the impact on packet reordering under different re-sequencing buffer availabilities is presented. Analytical relationships for variation of Reorder Density (RD) and Reorder Buffer occupancy Density (RBD), two measures of packet reordering, are provided as packet streams pass through re-sequencing buffers. Simulations driven by Internet measurement traces are used to verify the analytical results. The model provides an overall perspective of the impact of re-sequencing buffers and allows for buffer allocation to meet bounds related to packet reordering.

Large peer-to-peer systems exhibit the presence of communities based on user interests. Resources commonly shared within individual communities are in general relatively less popular and inconspicuous in the system-wide behavior. Hence, such communities are unable to benefit significantly from caching and replication that focus only on the most dominant queries. A community-based caching (CBC) solution that enhances both community-wide and system-wide lookup performance is proposed. CBC consists of a suboverlay formation scheme and a local-knowledge-based distributed caching (LKDC) algorithm. Suboverlays enable communities to forward queries through their members. While queries are forwarded, the LKDC algorithm causes members to identify and cache resources of interests to them, resulting in faster resolution of queries for popular resources within each community. Distributed local caching requires global information (e.g., hop count and popularity of contents) that is difficult and costly to obtain. However, by means of an analysis of globally optimal behavior and structural properties of the overlay, we developed the heuristic-based LKDC algorithm that not only relies on purely local information but also provides close-to-optimal caching performance. CBC is adaptive to changing popularity and user interests, works with any skewed distribution of queries, and introduces minimal modifications and overhead to the overlay network.

Emerging collaborative Peer-to-Peer (P2P) applications rely on resource discovery solutions to aggregate groups of heterogeneous, multi-attribute, and dynamic resources that are distributed. In the absence of data and understanding of real-life resource and query characteristics, design and evaluation of existing solutions have relied on many simplifying assumptions. We first present a summary of resource and query characteristics from PlanetLab. These characteristics are then used to evaluate fundamental design choices for multi-attribute resource discovery based on the cost of advertising/querying resources, index size, and load balancing. Simulation-based analysis indicates that the cost of advertising dynamic attributes is significant and in-creases with the number of attributes. Compared to uniform queries, real-world queries are relatively easier to resolve using unstructured, superpeer, and single-attribute dominated query based structured P2P solutions. However, they cause significant load balancing issues in all the designs where a few nodes are mainly involved in answering majority of queries and/or indexing resources. Moreover, cost of resource discovery in structured P2P systems is effectively O(N) as most range queries are less specific. Thus, many existing design choices are applicable only under specific conditions and their performances tend to degrade under realistic workloads.

A novel scheme is presented that allows individual nodes in sensor networks to achieve network/topology-awareness by listening to regular packets associated with applications. Nodes, initially oblivious to network topology and their position within the network, gradually infer information required to evaluate their own Virtual Coordinates (VCs). A Singular Value Decomposition based transformation allows each node to convert the VCs of nodes, gleaned from the source or destination address field of packets, to corresponding Topological Coordinates (TCs). Eventually each node generates a topology map of the network, thus becoming aware of its own location and those of other nodes. Effectiveness of self-learning scheme, in terms of convergence of different stages, and gradual development of network awareness at nodes are illustrated. While many applications of network awareness within nodes can be foreseen, we illustrate how performance of routing can improve dramatically over time as network awareness develops within a node.

Distributed, multi-attribute Resource Discovery (RD) is a fundamental requirement in collaborative Peer-to-Peer (P2P), grid, and cloud computing. We present an efficient and load balanced, P2P-based multi-attribute RD solution that consists of five heuristics, which can be executed independently and distributedly. First heuristic maintains a minimum number of nodes in a ring-like overlay consequently reducing the cost of resolving range queries. Second and third heuristics dynamically balance the key and query load by transferring keys to neighbors and by adding new neighbors when existing ones are insufficient. Last two heuristics, namely fragmentation and replication, form cliques of nodes that are placed orthogonal to the overlay ring to dynamically balance the highly skewed key and query loads while reducing the query cost. By applying these heuristics in the presented order, a RD solution that better responds to real-world resource and query characteristics is developed. Simulations using real workloads are used to demonstrate its efficacy.

Though resource discovery is a fundamental requirement in collaborative peer-to-peer, grid, and cloud computing, very little is known about resource/query characteristics and their impact on resource discovery. Fundamental design choices for distributed resource advertising and querying are evaluated in the context of existing practical systems. First, a generic model for cost of resource discovery is presented. Second, multi-attribute resource and query characteristics from Planet-Lab and SETI@home are presented. We observe that attributes of both resources and queries are highly skewed, correlated, queries are less specific, and Generalized Pareto distribution is suitable for capturing the distribution of most dynamic attributes and their rate of change. Based on these observations, different design choices are evaluated for resource discovery in terms of their cost of advertising/querying, latency, load balancing, and routing table size. The findings indicate that superpeer-based architectures have the potential to support large-scale resource aggregation as they simultaneously balance the cost and load.

Modeling and simulation of Peer-to-Peer (P2P) resources with correlated static and dynamic attributes is essential in application design, validation, and performance analysis. A novel mechanism is presented to generate realistic synthetic traces of multivariate static and dynamic attributes of P2P resources. The methodology is demonstrated using characteristics of PlanetLab node traces. First, a multi-attribute resource model is defined using a selected set of static and dynamic attributes. Second, characteristics of resources are presented. We observe that attribute values are correlated, follow a mixture of probability distributions, and time series of some of the dynamic attributes are nonstationary. Third, random vectors of static attributes are generated using empirical copulas that capture the entire dependence structure of multivariate distribution of attributes. Finally, time series of dynamic attributes are randomly drawn from a library of multivariate-time-series segments extracted from PlanetLab traces. These segments are identified by detecting the structural changes in time series corresponding to a selected attribute. Time series corresponding to rest of the attributes are split at the same breakpoints and randomly drawn together to preserve their contemporaneous correlation. Furthermore, a tool is developed to automate the synthetic data generation process and its output is validated using statistical tests.

A novel method of identifying boundaries of wireless sensor networks deployed on 2D and 3D surfaces is presented. It does not require costly, error prone localization algorithms or physical locations of nodes. Instead, a Virtual Coordinate System (VCS) is used in which each node is characterized by the hop- distances to a set of randomly selected nodes known as anchors. To use geometric relationships for boundary detection, it transforms the VCS to a Topology Preserving Map (TPM). A TPM generation scheme for networks deployed on 3D surfaces is derived as well. The boundary detection scheme proposed is simple, not computationally intensive, energy efficient, and can be used with physical coordinates as well. Five representative example networks show the proposed scheme to be effective, with 100% of boundary nodes identified correctly with no erroneous identification of non-boundary nodes as boundary nodes. Use of TPM based boundary detection scheme for detecting dynamic event boundaries, such as those of plumes, in a distributed manner is also illustrated.

Large Peer-to-Peer (P2P) systems for file transfer exhibit the presence of communities based on semantic, geographic, or organizational interests of users. Generally, resources commonly shared within individual communities are relatively unpopular and inconspicuous in the system-wide behavior. These communities are unable to benefit significantly from performance enhancement schemes such as caching that focus only on the most dominant queries. We propose a generic caching framework that enhances lookup performance of individual communities while providing even better performance to the dominant communities. The caching framework can be used with any structured P2P system that provides alternative paths to a given destination. Furthermore, the solution is adaptive to changing popularity and user interests, works with any skewed distribution of queries, needs small caches, utilizes local statistics, and introduces minimal modifications and overhead to the overlay network. Simulations based on Chord overlay show 40% reduction in average path length with individual communities indicating three times improvement in performance over system-wide caching.

A Directional Virtual Coordinate System (DVCS) is proposed based on a novel transformation that restores the lost directionality information in a Virtual Coordinate System (VCS). VCS is an attractive option to characterize the node locations in Wireless Sensor Networks (WSNs), instead of using geographical coordinates, which is expensive or difficult to obtain. A VCS characterizes each node in a network with the minimum hop distances to a set of anchor nodes as its coordinates. The proposed transformation supplements the virtual coordinates, thus preserving all the inheriting properties such as embedded information of geodesic distances in the coordinates. The virtual directionality introduced, alleviates the local minima issue present in original VCS. Properties of this virtual directional domain are discussed. With these directional properties, it is possible, for the first time, to consider deterministic algorithms in the virtual domain, as illustrated with a constrained tree network example. A novel routing scheme called Directional Virtual Coordinate Routing (DVCR), which illustrates the effectiveness of the Directional Virtual Coordinate domain is proposed. DVCR significantly outperforms existing VCS routing schemes Convex Subspace Routing (CSR) and Logical Coordinate Routing (LCR), while achieving a performance similar to the geographical routing scheme - Greedy Perimeter Stateless Routing (GPSR), but without the need for node location information.

Robust BaseLine (RBL) is a formal technique for extracting the baseline of network traffic to capture the underlying traffic trend. A range of applications such as anomaly detection and load balancing rely on baseline estimation. Once the fundamental period of the pattern for analysis is recognized, e.g., based on user interest or a period detector such as Autocorrelation Function (ACF), the basic extraction is carried out in two steps. First, the common component across the dataset is separated using Robust Principal Component Analysis (RPCA). The fundamental pattern in the common component is extracted using Principal Component Analysis (PCA) in the second step. Scaling factors required to fit the base-pattern back into the data are returned automatically by PCA. Two types of traffic baselines may be extracted: RBL-L captures the common behavior across time on a single link, and RBL-N captures the common behavior across a network of links, i.e., in space. RBL-N is particularly useful for specifying traffic matrices more efficiently over time, which normally requires multiple updates to follow baseline trends. The derived base-patterns for a single link or a single time period is then extended over the entire network or thru the entire observation period with a compressive analysis. The compressed base-pattern provides a smoother baseline and also a filter to separate baseline traffic and the deviations on the fly from traffic measurements. When compared against BLGBA (Baseline for Automatic Backbone Management) the proposed scheme provides a less noisy, more precisely fitting baseline. It is also more effective in revealing anomalies.

Virtual Coordinate Systems (VCS) characterize each node in a network by its hop distances to a subset of nodes called anchors. Performance of VCS based algorithms is highly sensitive to number of anchors and their placement. Extreme Node Search (ENS), a novel and efficient anchor placement scheme, is proposed that demonstrates significantly improved performance over state-of-the-art. ENS starts with two randomly placed anchors and then uses a Directional Virtual Coordinate (DVC) transformation, which restores the lost directionality in traditional VCS, to identify anchor candidates in a completely distributed manner. A vector-based representation is proposed for the DVC domain, which is then used to introduce the concept of angles between virtual directions in the transformed domain. Ability to specify cardinal directions and use angles is a radical change from the traditional VC system approaches. By selecting two anchor pairs with near orthogonal directional coordinates under the DVC transformation, a novel Topology Preserving Map (TPM) generation scheme is developed. This new TPM generation scheme requires significantly less computations than the existing PCA based method. Use of ENS significantly enhances the PCA based TPM generation as well. Simulation results for representative WSNs indicate that the ENS based anchor sets significantly improve performance of all prominent VC based routing schemes. For instance, Directional Virtual Coordinate Routing, when combined with ENS anchor placement strategy, outperforms even the geographic Greedy Perimeter Stateless Routing scheme that relies on exact physical node coordinates.

Geo-Logical Routing (GLR) is a novel technique that brings the advantages of geographic routing to logical domain, without inheriting the disadvantages of physical domain, to achieve higher routability at a lower cost. It uses topology domain coordinates, derived solely from virtual coordinates (VCs), a better alternative for location information. In logical domain, a node is characterized by a VC vector, consisting of minimum number of hops to a set of anchor nodes. VCs contain information derived from connectivity of the network, but lack physical layout information such as directionality and geographic voids. Disadvantages of geographic routing, which relies on physical location information, include cost of node localization or/and use of GPS, as well as misrouting due to physical voids. With the ability to generate topological maps from virtual coordinates via a Singular Value Decomposition based technique, it is now possible to characterize a network with topological coordinates, which are shown to be more effective than physical coordinates for making routing decisions. By switching between a geographic routing scheme operating on topological coordinates and a logical routing scheme, GLR overcomes local minima in the respective domains. Performance results presented indicate that GLR significantly outperforms existing logical routing schemes - Convex Subspace Routing (CSR) and Logical Coordinate Routing (LCR) - as well as geographic scheme, Greedy Perimeter Stateless Routing (GPSR).

Virtual Coordinate System (VCS) based routing schemes for sensor networks characterize each node by a coordinate vector of size M, consisting of distances to each of a set of M anchors. Higher the number of anchors, the higher the coordinate generation cost as well as the communication cost. Identifying an effective set of anchors and encapsulating original VCS's information in a lower dimensional VCS will enhance the energy efficiency. Two main contributions toward this goal are presented. First is a method for evaluating the amount of novel information contained in an ordinate, i.e., in an anchor, on the coordinate space created by the rest of the anchors. This method can be used to identify unnecessary or inefficient anchors as well as good anchor locations, and thus help lower overhead and power consumption in routing. Second, a method for reducing the VCS dimensionality is presented. This Singular Value Decomposition (SVD) based method preserves the routability achieved in original coordinate space but with lower dimensions. Centralized and online realizations of the proposed algorithm are explained. Examples of different topologies with 40 anchors used in performance analysis show that coordinate length can be reduced on average by a factor of 8 without degrading the routability. Use of novelty filtering to select effective anchors prior to SVD based compression results in further improvement in routability.

Capacitive Leadframe Testing is capable of detecting open solder defects in Printed Circuit Boards (PCB). Principal Component Analysis (PCA)-based approach has been shown to be effective in identifying outlier devices using Capacitive Leadframe Testing measurements. In practice, when a sense plate orientation is shifted or tilted, the resulting measurement variation makes detecting outliers harder. Approaches are introduced to compensate for the `abnormal' measurements due to sense-plate variations. A PCA based technique is developed to estimate the relative amount of tilt and shift in sense plates. Such estimates can be used to compensate for mechanical misalignments. It can also isolate the misalignment related information from the defect related information in the data. The effectiveness of this technique in the presence of the two common forms of mechanical variations is illustrated using experimental measurements from a laboratory setting. The approach is not sensitive to the order of pins, and as such, shows promise for detection of complex but systematic errors introduced by sense plate misalignments.

A new approach based on graph wavelets for analyzing the spatial and temporal behavior of Internet traffic anomalies is presented. This approach is applied to Internet2 traffic measurements to evaluate the time duration and spatial spread (number of links affected) of anomalies. Based on the empirical results, a node model is proposed that captures the behavior of anomalies at individual network nodes. The model considers various aspects of anomalies, such as its origin, termination, propagation, duration and volume changes. The derivation of the model parameters requires only local node information, but the model is capable of producing network-wide anomalies whose behavior mimics network wide anomalies. Model is verified by using Internet2 traffic data. Since the proposed model can be specified using only a few parameters, it can be used in place of large anomaly traces with a great data reduction. As extensions, the model is applied over a path and an aggregated model that applies to a neighborhood in the network is also presented. A method to use the graph wavelet components found during the analysis to implement a real-time anomaly monitoring system is also discussed.

A method of obtaining a topology preserving map from virtual coordinates of a sensor network is presented. In a Virtual Coordinate System (VCS), a node is identified by a vector containing its distances, in hops, to a set of nodes called anchors. VCS is a higher dimensional abstraction of the connectivity map of nodes, with dimensionality defined by the number of anchors. Physical layout information such as physical voids and even relative physical positions of sensor nodes with respect to X-Y directions are absent in a VCS description, and obtaining the physical topology has not been possible up to now. A novel technique, based on Singular Value Decomposition, is presented to extract a topology preserving map from VCS. Three options with different computation and communication complexities, as a result of using different subsets of coordinates as the input, are presented and analyzed; the input for the three cases consist of a) the entire virtual coordinate set, b)only the virtual coordinates of anchors, and c) virtual coordinates of a random set of nodes. Evaluation results indicate that last two approaches achieve comparable accuracy to the first, but with significantly less complexity. Topology preserving maps for networks representing a variety of topologies and shapes are extracted. A new metric termed Topology Preservation Error (ETP) is defined to evaluate the topology preservation; it accounts for both the number of node flips and degree of the flips. The techniques extract topology preserving maps with ETPless than 2%

Random routing protocols in sensor networks forward packets to randomly selected neighbors. These packets are agents carrying information about events, or queries seeking such information. We derive the probability of a packet visiting a given node in a given step as well as the rendezvous probability of agents and queries within a specific number of hops at a given node(s) in a 2-D grid-based sensor network. The utility of the model is demonstrated by determining the protocol parameters to optimize performance of rumor routing protocol under different constraints, e.g., to evaluate the number of queries and agents to maximize the probability of rendezvous for a given amount of energy. Monte Carlo simulations are used to validate the model. The closed form exact solution presented, unlike existing models relying on asymptotic behavior, is applicable to small and medium-scale networks as well. An upper bound is provided for the case where the packet is not sent back to its immediate forwarding node. Simulation results indicate that the model is a good approximation even for sparse arrays with 75 % of the nodes. The model can be used to set parameters and optimize performance of several classes of random routing protocols.

Capacitive Leadframe testing is an effective approach for detecting faults in printed circuit boards. Capacitance measurements, however, are affected by mechanical variations during testing and by tolerances of electrical parameters of components, making it difficult to use threshold based techniques for defect detection. A novel approach is presented for identifying boards that are likely to be outliers. Based on Principal Components Analysis (PCA), this approach treats the set of capacitance measurements of individual connectors or sockets in a holistic manner to overcome the measurement and component parameter variations inherent in test data. The effectiveness of the method is evaluated using measurements on three different boards. Enhancements to the technique to increase the resolution of the method are presented and evaluated.

Virtual coordinate based routing (VCR) is a class of simple scalable routing schemes for sensor networks. VCR relies on coordinates based on the number of hops to a set of anchors, rather than on geographical coordinates. Multiple nodes with identical coordinates and local minima encountered during routing degrade the performance of VCR. Properties of virtual coordinate systems are analyzed to provide insight into the nature of virtual coordinate space, and bounds are derived for path lengths. A new routing scheme, Convex Subspace Routing (CSR), is proposed. In contrast to existing VCR schemes that use backtracking or hill climbing techniques to overcome local minima, CSR avoids using anchors that cause local minima. CSR selects subsets of anchors dynamically to provide a convex distance function from source to destination. Consequently, it is less sensitive to anchor placement and over anchoring, and does not require tracking of route history for backtracking. Performance of CSR is evaluated for different network configurations. Results for varying node densities and different numbers of anchors indicate that CSR significantly outperforms Logical Coordinate Routing (LCR), in terms of routability, number of hops used for routing, memory usage as well as energy consumption.

Multi-sensor data fusion latency in Internet-based sensor applications is analyzed. The probabilistic estimation of periodic backlog (PEPB) technique presented takes into account the nature of sensor data generation and the time-scale invariant burstiness (i.e., self-similarity) of network traffic. The fusion application considered requires synchronizing a set of correlated data before fusion processing can begin. PEPB is used first to model the single-hop delay. Analysis is then extended to a network with multi-hop communication and multiple sensor nodes. Comparison with simulation-based results demonstrates the accuracy of the model.

Welcome to the Fourth IEEE LCN Workshop on Network Measurements.

The first generation testbed of the Center for Collaborative Adaptive Sensing of the Atmosphere (CASA-IP1) is currently operational in Oklahoma. The CASA-IP1 system is a radar network observing a weather event simultaneously by four radars. Within the CASA, a network-based reflectivity retrieval technique has been developed and evaluated extensively. This paper presents the evaluation of the network-based reflectivity retrieval by comparing retrieval results using CASA-IP1 data with WSR-88D observations. This paper also describes the design and implementation of an architectural framework for real-time processing of the network-based retrieval algorithm. Experimental results show that the implementation meets the real-time requirement of CASA.

A closed-loop contaminant plume monitoring system is being developed that integrates wireless sensor network based monitoring with numerical models for subsurface plumes. The system is based on a novel virtual sensor network architecture that supports the formation, usage, adaptation, and maintenance of dynamic subsets of nodes. This automated monitoring system is intended to capture transient plumes to assess the source, track plumes in real-time, and predict future plumes behavior using numerical models that can be continuously re-calibrated by sensor data. This paper presents recent progress made in (1) developing a proof-of-concept study using a porous media test bed with sensors deployed; and (2) developing distributed algorithms for virtual sensor networking.

Security is a prime concern in large-scale wireless sensor networks used for collaborative mission critical applications. A backbone network in the form of a cluster tree can enhance upper layer functions such as routing, broadcasting, in-network query processing and network management. A secure backbone based on the cluster tree enables secure upper layer functions and dynamic distribution of cryptographic keys among different nodes and users of collaborative networks. A secure cluster tree formation algorithm is presented that is independent of key pre-distribution scheme, network topology, and does not require a-priori neighborhood information or location awareness. Simulation based results show that the algorithm retains most of the desirable cluster and cluster tree characteristics while building the secure cluster tree. Availability of neighborhood information further improves the performance. Our simulations also suggest that hierarchical networks are more vulnerable to node capture than non-hierarchical networks.

A network-based reflectivity retrieval technique has been developed within the Center for Collaborative Adaptive Sensing of the Atmosphere (CASA). The concept of a networked- radar system is simultaneous observations of the same precipitation event by multiple radars operating at the attenuating frequency such as X-band and scanning in a low elevation plane. This paper presents the preliminary demonstration of the network-based retrieval using data from the first Integration Project (IP1) radar network in Oklahoma. Electromagnetic waves backscattered from a common volume in a networked radar system are attenuated differently along the different paths. The CASA networked-retrieval method is based on a set of governing integral equations describing the backscatter and propagation of common volume with constraints of total path attenuation. The method has been implemented in a multiprocessor environment, which operate simultaneously and collaboratively to meet the real time requirement of CASA. The performance of the implemented retrieval algorithm such as computation requirement will be presented. Comparison of the CASA networked retrieval is made against the conventional attenuation correction based on the principle of coupling the specific attenuation, differential propagation phase and reflectivity. The preliminary results show good agreement with conventional differential phase base attenuation correction.

The network link speeds increase at a higher rate compared to processing speeds. This coupled with the increase in size of router tables demand higher levels of parallelism within router hardware. However, such parallelism introduces unintended consequences that potentially may negate some of the performance gains provided by the improved technology. The growth trends of computing speeds, link speeds, and routing table sizes are used to evaluate one such consequence, packet reordering within routers. Results presented show the trends related to the degree of hardware parallelism and packet reordering.

As sensor nodes are subject to strict power limitations and often deployed in harsh environments, there is significant potential for a node to be hidden from another node. The impact of hidden nodes on performance of the IEEE 802.15.4 low-rate wireless personal area network protocol is evaluated. At lower Tx power, as a result of hidden nodes, there are more collisions in the network increasing the cost of packet delivery. The simulation results indicate that there exists an optimum transmission power that minimizes the impact of hidden nodes, and it depends on factors such as network load, desired throughput and beacon order. Results indicate that additional throughput gain and lower cost per packet delivered can be achieved by increasing the beacon order at higher Tx power.

An approach is presented to capture and model Internet end-to-end packet delay behavior using ARMA and ARIMA models. Autocorrelation (ACF) and Partial Autocorrelation (PACF) functions are used to identify the most appropriate model and the model order. Impact due to sending rate and packet size of the probe, and the available link capacity on these two metrics are investigated. Results indicate that the models presented reflect accurately the effect of packet correlation induced by the network. Modeling Inter-Packet Gap (IPG) is an alternative for capturing the effect of the network on a packet stream. A methodology for fitting ARMA and ARIMA models to end-to-end packet delay and IPG series is presented.

Increased parallelism in routers necessary to handle high link speeds and large routing tables, wireless ad hoc routing, QoS provisioning, and overlay routing, are some of the factors that lead to an increase in reordering on the Internet. Packet reordering due to packet forwarding over multiple paths is investigated. An analytical model is derived for load splitting scenarios and verified using emulated topologies. The resulting reordering is profiled using reorder density, and analyzed with respect to path delays, path probabilities and number of paths. The variation of packet displacement with delay variation and forwarding probabilities is quantified. The special case corresponding to two paths is evaluated in detail. For any load splitting, the increase in the difference in the delay between paths leads to increased reordering, making the paths with closer delay values more preferable. The model can also be applied to a single-path case where reordering is caused by wide delay variation among packets, by deriving an equivalent set of probabilities corresponding to path splitting scenario.

In many emerging sensing applications, random network losses may lead to drop of critical information, rendering partially received data useless for the end users. A packet-marking scheme based on the application content is proposed that enables application-aware processing of the data within the overlay network. A token-bucket based rate control algorithm in conjunction with the proposed packet-marking scheme enables on-the-fly selection of data for forwarding/drop to a particular end user at the desired transmission rate. We demonstrate the effectiveness of the packet-marking and token-bucket based rate control scheme in simultaneously meeting heterogeneous QoS requirements of the multiple end users for content quality and bandwidth for a weather monitoring sensing application.

A statistical technique X-IDDQ for extracting defect information from IDDQ data is presented that is effective for detection of defects in ICs. The technique treats the IDDQ measurements in a holistic manner to come up with a statistic X that is highly correlated to the presence of defects. X-IDDQ facilitates binning of ICs and enhances the test process by early identification of faults. The transformation metrics, for evaluating X statistic from IDDQ measurements, obtained using one batch works extremely well for different batches, facilitating its use with manufacturing-line testing.

A performance model is presented for an optical packet switch architecture in which the wavelength converters are shared per output link and each output link consists of multiple fibers. Symmetry of the switch is exploited to derive the packet loss probability for the case where traffic is destined to different output ports with equal probability. The architecture performance is evaluated by means of an analytical model and confirmed by simulations under different switch parameter configurations. Wavelength converters are shown to improve the packet loss probability of the switch. The study shows that synchronous switches equipped with full conversion would have the least conversion utilization rate indicating that the use of a switch with less converter count, i.e., partial conversion, would offer better switch resources utilization and comparable packet loss performance

A formal approach for characterizing, evaluating and modeling packet reordering is presented. Reordering is, a phenomenon that is likely to become increasingly common on the Internet, and it can have an adverse impact on end-to-end performance and network resource utilization. Reorder density (RD) is a comprehensive metric for reordering. Several properties of RD are presented that provide insight into the nature of reordering present in a sequence. Based on measurements of packet reordering, certain basic patterns of reordering that are prevalent on the Internet are identified. By focusing on these basic patterns, a model is developed for reordering in networks involving two parallel paths, using a simple load-balancing scenario as an example. The model is verified using an emulation testbed

Network survivability gauges the ability of a network to support the committed quality of service (QoS) continuously in the presence of various failure scenarios. We perceive that both performance and availability are integral components of survivability. Therefore, we propose a composite model for survivability that consists of performance and availability analysis. A analytical technique is presented to find the excess loss due to failure (ELF) when the system is operating in gracefully degraded states. An algorithm is proposed to carry out the availability analysis of the network, even when the available paths between nodes are non-disjoint. These two models are combined to construct a hierarchical model to evaluate the network survivability performance. We consider single and multiple link failures, and also node failures. A WDM network with wavelength conversion is used as an example for this evaluation.

Gigabit networking makes possible the remote access to expensive and specialized facilities that were inaccessible in the past due to limited bandwidths. VCHILL project for digitized radar data transfer is one such real-time application that will tap the next generation Internet technology to provide interactive access to real-time and stored data generated by weather radars, thus revolutionizing the way experiments are carried out. The design and implementation of the deployment of an efficient congestion control algorithm for this application are presented. The TCP-friendly rate adaptation based on loss (TRABOL) algorithm is a source-based rate control mechanism that controls the transmission rate based on the feedback about losses experienced by the client station. The performance results show that the deployment of this algorithm makes the application TCP-friendly.

The tradeoffs involving the use of multifiber multihop networks such as the number of fibers, number of wavelengths, conversion options, and the different switch configurations are examined, and their impact on end-to-end blocking and throughput performance is evaluated. Models relating network parameters to end-to-end performance of circuit switched all-optical networks are developed. The performance gain due to the use of multiple fibers with or without conversion proved to be superior to the single fiber case, when the total number of wavelength are the same, for traffic that traverses multiple hops between source and destination nodes.

Implementing a TCP-friendly congestion control mechanism is imperative for emerging UDP-based real-time, high-bandwidth applications such as multimedia. VCHILL project for digitized radar data transfer over the Internet is one such application for which a source based TCP-friendly rate adaptation based on loss (TRABOL) algorithm is deployed and shown to be TCP-friendly over applicable timescales. Comparison of the TRABOL-based radar application with other non-congestion controlled UDP flows shows that the application performs better with TRABOL and is fair towards neighboring TCP flows. Implementation of a simple memory-based mechanism as an add-on to the TRABOL algorithm enhances the performance of the application.

The performance of an optical switch that handles contention resolution using a fiber delay line (FDL) is modeled and evaluated. We propose using a simple optical buffer that contains only a single FDL for contention resolution. The analytical model derived for the performance of an optical switch based on this buffer is verified using simulations. The analytical model can be utilized with both packet and burst switching schemes to characterize the performance of switches augmented with this FDL architecture.

The accuracy of the currently available bottleneck capacity measurement tools that use the packet pair technique is significantly affected by cross traffic and queuing delays. The impact of these factors is investigated, and a new technique is presented to overcome these effects by estimating the jitter based on packets sent around the same time as the packet pair. A tool based on this principle, Captool, measures the end-to-end bottleneck capacity by reducing these adverse effects, thereby increasing the accuracy of the end-to-end measurements.

Today's ISP's are being obligated to rely on architectures capable of classifying incoming streams based not only on their demand requirements but also on the willingness of the user to pay. We evaluate three Internet pricing models that along with an accurate QoS architecture, are intended to constitute a complete mechanism capable of rapidly reacting against the user response and willingness to pay, and thus manage the network resources intelligently.

Out of order arrival of packets is an inevitable phenomenon on the Internet. Application performance can degrade to a great extent due to out-of-order arrival of packets. Metrics to characterize the degree of re-ordering will promote the evaluation of network protocols with respect to packet reordering as well as provide a uniformly defined quantitative measure for packet reordering. We define the concept of "reorder density function" (RD), which quantitatively measures the degree of reordering in sequences of packets. While the reorder density function can provide a comprehensive picture on reordering, measures such as mean and median of reorder density can be used when simpler metrics are desired. Results of reordering analysis using the above metrics for TCP and UDP based applications under different network conditions are presented.

Differentiated Services (DiffServ) is an important mechanism for providing quality of support on Internet. Even though DiffServ has demonstrated its stream classifications ability, the lack of an efficient network resource management mechanism is still a dilemma for this architecture. To address the mentioned problem, this paper proposes two new bandwidth brokering approaches that by using the network availability information along with the DiffServ priority discipline accomplish an efficient dynamic stream allocation mechanism within the DS domain (DiffServ domain). They are able to address intelligently the network resources based on the users priorities and the application requirements. These two new approaches use the DifServ class-based information provided by the queueing mechanism in order to estimate the link congestion for each stream type. By analyzing this data our mechanisms can choose the suitable path that best satisfies the user requirements and privileges. The results presented successfully demonstrate the effectiveness of the proposed mechanisms and their advantages over other existing approaches.

The problem of reconfiguring wavelength-routed optical networks when the traffic demand changes is considered. When reconfiguring the wavelength assignment according to the traffic demand, it is important to reduce the number of receivers that need to be retuned. Furthermore, after the reconfiguration, the load needs to be balanced among the wavelengths. The tradeoff here is between the number of retunings and the degree of network load balance. After tracking real traffic demand to see the effect of traffic changes on the network load balance, an algorithm has been developed for wavelength reconfiguration, so that the demand on the most loaded channel is reduced by exchanging one node with the least loaded channel. The most and least loaded channel balance (MLLCB) algorithm provides very good results in terms of the required number of retunings and the load balance. Comparison of MLLCB algorithm with GLPT algorithm shows a very significant improvement in performance.

Multihop networks such as the Manhattan Street Network (MSN) are strong contenders for implementing high-speed networks, ranging from back-plane networks for cluster computing to metropolitan-area networks. Optical buffering may be used to augment optical multihop networks, and to reduce or even eliminate the need for optical-electrical conversions. This paper evaluates the performance enhancements obtainable by augmenting traditional MSN with optical buffering, thus combining optical buffering and deflection routing for contention resolution. The network is based on a novel 6/spl times/6 optical switching node. A great improvement in performance over the traditional MSN is obtained for a large range of network parameters.

Effectiveness of the clustering based approach in detecting devices with abnormal I/sub DDQ/ values is evaluated using data from the SEMATECH test methods experiment. The results from clustering are compared to the results obtained on actual silicon during the SEMATECH study. The differences between the results obtained in each case are analyzed. The clustering approach is also compared to two common I/sub DDQ/ test techniques, the single-threshold approach and the delta-I/sub DDQ/ approach, and the results are presented.

A new technique for evaluating I/sub DDQ/ data using a clustering based approach is presented. While prevailing I/sub DDQ/ test techniques rely on a fixed threshold or the current signature of an IC, the proposed technique relies on abnormalities of the I/sub DDQ/ distribution of a device with respect to other devices in the test set. Results of applying this technique to data collected on a high volume graphics chip are described. Results are also compared to the conventional single threshold approach, and benefits of the new technique are presented.

The robust-WDM is proposed to realize wavelength division multiplexed (WDM) local area networks (LANs) in the presence of laser wavelength drifts. The approach of robust-WDM is to design access protocols that can exploit a new wavelength-tracking receiver. We introduce an analytical model that can be used as a tool to design robust-WDM LANs, and to predict and understand their performance characteristics. The model is designated for a class of these networks that relies on an aperiodic-reservation based access protocol and uses a token-passing control channel to organize the right to reserve among nodes.

This paper introduces the concept of antirandom testing where each test applied is chosen such that its total distance from all previous tests is maximum. This spans the test vector space to the maximum extent possible for a given number of vectors. This strategy results in a higher fault coverage when the number of vectors that are applied is limited. Results on several ISCAS benchmarks show this strategy to be very effective when a high fault coverage needs to be achieved with a limited number of test vectors. The superiority of the antirandom testing approach is even more significant for testing bridging faults.

The hidden terminal problem is unique to wireless networks and as of now there is very limited understanding about its effects on network performance. Results are presented from a simulation study of the IEEE 802.11 MAC protocol when operating in the presence of hidden terminals. We also propose a framework for modeling hidden terminals which can handle complex scenarios of both mobility and static obstructions. Our simulations indicate that hidden terminals can have a very detrimental effect on the performance of the IEEE 802. 11 MAC protocol. Although the throughput is acceptable when about 10 percent of station pairs are hidden, the packet delay can increase by an order of magnitude. The performance of the protocol drops sharply when the number of hidden pairs exceeds 10 percent.

Current communication subsystem mechanisms within workstation and PC class computers are limiting network communication throughput to a percentage of the present network data rates. Though CPU and computer network speeds increased by more than an order of magnitude, the execution rate of computer functions and applications requiring network communications have increased only marginally. An integrated network communications architecture (INCA) is presented that is interoperable with all existing programs, computers and networks, and scales with network and CPU speeds. The INCA architecture minimizes internal system limitations and provides application level, as opposed to network interface level, internal machine data throughput rates near those of high speed network transmission rates. Test results of a software implementation of the INCA architecture on actual systems and networks are presented that show a 260% to 760% improvement in the application level throughput of network communicated data of workstation and PC class computers.

A number of research issues are considered in the study of an European multiwavelength optical network. The main part of the study is the design of an all-optical network to overlay the European national networks and link major centers therein. We survey most of the issues considered in this respect within the frame of two projects funded by the European Commission. Another part of the study focuses on the underlying European national networks and their possible development to interface with the proposed overlay. The main aim of this paper is to put together these two parts and oversee the prospect of an European ultra-high capacity multiwavelength optical infrastructure.

Many WDM architectures are based on fixed wavelength channels which require the lasers to be tuned to their channels accurately. Achieving this task is difficult and costly especially in the distributed environments. Robust WDM is an approach that tolerates large laser wavelength variations, due to temperature drifts and manufacturing tolerances. A reservation based medium access protocol is used to dynamically select the laser for communication. A token passing based control channel assigns a reservation interval to one of the waiting stations. The performance of a circuit switched robust WDM network is considered when each node has only a limited number of lasers. The model evaluates the performance of a network for a given access pattern, specifying the channels accessible to each station. Simulation results are used to verify the analytic results.

Combining the advantages of bipolar and CMOS, BiCMOS is emerging as a major technology for high speed, high performance, digital and mixed signal applications. Recent investigations have revealed that bridging faults can be a major failure mode in ICs. This paper presents the effects of bridging faults affecting p- or n-parts and input bridging faults of logical nodes affecting p- and n-parts. It is shown that bridging faults can be detected by I/sub DDQ/ monitoring in BiCMOS devices. An input pattern classification scheme is presented for bridging faults. These classes of input patterns are then used to obtain test sets for bridging fault detection.

Looking toward the future of computer networking, we see that the need for increased interconnectivity will require continual growth in speed and features. Today there are several successful LANs competing in the 100 Mbps arena. Now we are considering the next generation of network. Ideally, users need a solution that provides the necessary bandwidth and connectivity while preserving their investments in hardware and software. This paper will present an architecturally different design for a new more effective LAN. The new architecture will provide a convenient solution to the need for higher total network capacity by interconnecting multiple FDDI networks (and/or network segments). The underlying network architecture is based on the FDDI protocols, which is an attractive feature for those who have invested resources in FDDI. Scalability is a required feature for any future network design and is addressed in detail by this architecture. This alternative next-generation networking solution has been presented to the ANSI standards committee X3T12 as a possible FDDI follow-on standard (FFOL).

The performance of a symmetric robust WDM network that uses token-passing on the signaling channel and has a limited number of lasers per station is modeled and analyzed for circuit-switched connections. Robust WDM networks use a reservation interval during which all stations cease their WDM transmissions so that the station holding the token may establish a connection with its intended destination. The network performance is analyzed for different network parameters. Simulation results are used to evaluate the accuracy of the model.

This paper explores the effects of locality on the performance of a FDDI-based multinetworking environment and compares this to a scaled FDDI approach. Both of these schemes have been presented to the Standards Committee for consideration in migrating the FDDI protocols to gigabit speeds. One is a simple linear scaling (10/spl times/) of the FDDI protocol to reach the gigabit speed, and the second, presented by these authors, is a hierarchical scheme which preserves the existing FDDI standards and defines an interconnection network that acts as a backbone to the underlying FDDI networks. An introduction to both network schemes is presented. The average network access delay in the hierarchical network scheme is affected by the locality of the data traffic since data traffic has to traverse two networks if it is not local. A numerical method is presented which allows access to delay results with various network configurations. In the scaled approach there is only one network, thus eliminating the effects of data locality on the performance of the network. The paper is concluded with a discussion of the pros and cons of the two schemes.

Artificial neural networks provide an attractive alternative in performing the policing function at the user network interface (UNI) of an asynchronous transfer mode (ATM) network. In order to guarantee quality of service (QOS) for the established connections in ATM networks, one of the policing functions at the UNI is to ensure that all data streams entering the ATM network conform to the allocated bandwidth, or otherwise the cell loss priority (CLP) bit in the ATM cell header must be set to reflect the situation that the output of the UNI has exceeded the permissible bandwidth. Feed-forward neural networks with back-propagation learning algorithms are chosen to perform the policing function at the UNI. Numerical results are presented to illustrate that the neural network is capable of performing the policing function.

The classical M/G/1 system with vacations has been used to estimate the mean waiting time of individual priority classes of homogeneous FDDI networks. To faithfully model the FDDI network using M/G/1 system, the vacation interval of the server (token) as seen by individual priority classes must be evaluated. The expressions for the mean waiting time of individual priority classes are functions of first and second moments of the vacation intervals of that class. It can be shown that the expression derived for the first and second moments of the vacation intervals of the respective classes are functions of the mean and variance of number of messages transmitted by each priority class per token rotation. These expressions are evaluated using an iterative procedure that utilizes the prior knowledge of the throughput characteristics of the FDDI network. Simulation results show that the model provides accurate estimates for the mean waiting time of individual priority classes of a homogeneous FDDI network.

A new BiCMOS dynamic logic family is presented. The logic gates provide a significant speed-up over existing logic families, such as, CMOS, BiCMOS and dynamic CMOS for the same feature size. The proposed logic family provides high drive capability to drive large loads at higher speeds compared to CMOS domino logic family; there are power advantages due to dynamic operation compared to the conventional fully complementary CMOS and BiCMOS devices. It also has area advantage compared to that of the conventional BiCMOS gates. The proposed high-speed BiCMOS domino family provides speed improvement with existing feature size, using the existing fabrication technology.

WDM networks in general depend on transmitters and receivers that are precisely tuned to predetermined fixed wavelengths. Robust WDM is an approach aimed at relaxing manufacturing and operating wavelength tolerance requirements, which will lead to cost effective implementations. In this approach, connections are based on dynamically selected wavelengths as opposed to using fixed wavelength channels. A station that needs a connection selects a wavelength during a reservation interval. The performance of a simplified WDM protocol with dynamic wavelength selection is modeled and analyzed for packet traffic. In this version, the reservation intervals are assigned to stations randomly. The performance of the protocol is analyzed for different network parameters.

Simulators are valuable tools in helping us understand how complex systems will operate and perform, without actually having to build a physical model to test. It is best to write the simulator in a higher level programming language, in our case we use Mathematica and are able to obtain fast results with excellent graphing capabilities. The goal of our research is to find a method for obtaining an estimate of the average access delay for advanced token ring based networks such as FDDI (fiber distributed data interface) or FFOL (FDDI follow on LAN). We have developed a method that will generate exact results at low to medium loaded networks. As the load level increases, the error rate also increases, this however is monitored and can be used as a confidence measure for the results obtained. A symmetric multimedia environment is used as the model, with bimodal traffic consisting of short data packets and longer video packets. The method we developed to calculate the estimated average access time under a specific network configuration as a function of varied offered load. A full listing of the Mathematica code is included in the appendix and is referenced in the analysis section.

Multichannel optical networks promise bandwidth in the gigabits/sec domain. This current trend in optical device technology demands a simple yet scalable protocol for future high speed networks. The Request Based Channel Access (RBCA) protocol provides solution based on folded bus topology. RBCA protocol communication consists of two steps, namely, bandwidth request and data transmission. The communication is slot based and supports variable packet lengths. A dedicated monitor node in the network performs all protocol processing and error detection/correction functions, and minimizes all protocol overheads in nodes. The protocol provides fair channel access to all nodes, irrespective of their relative position from the fold of the network. In a multichannel network, the communication of each channel is independent of all other channels. Thus multichannel networks can be implemented as better optical communication devices, supporting large number of parallel channels, become available.

The waiting time and queue length characteristics of isochronous, synchronous and asynchronous traffic at the gateway between FDDI-II and ATM networks are analyzed. A generalized approach to analyze various classes of traffic separately, is discussed. We first present an overview of our gateway model. Next, a deterministic analysis of isochronous traffic is presented. Synchronous and asynchronous classes of traffic are analyzed with the gateway model used. These results are compared with those generated by our simulator.

Local area networks (LANs) have emerged as the technology of choice for inter-computer communication. As computers get faster, so must the LAN environment that supports them. Today there are several successful competing LANs in the 100 Mbps arena, and we are now starting to look forward to the next generation of network to take over when the current generation cannot provide the necessary bandwidth and connectivity. The paper discusses some of the architectural trade-offs available to designers of the next generation LAN, and proposes an architecture which provides for convenient interconnection of multiple FDDI networks (and network segments). The underlying network architecture is based on FDDI protocols, an attractive feature for those who have invested into FDDI. Scaleability appears to be a required feature for any future network design and is addressed in detail. The ultimate goal of this research is to present an alternative next generation networking solution that could be adopted by ANSI as the FDDI follow on standard.>

In BiCMOS, transistor stuck-OPEN faults exhibit delay faults in addition to sequential behavior. Stuck-ON faults cause enhanced I/sub DDQ/. The faulty behavior of Bipolar (TTL) and CMOS logic families is compared with BiCMOS. The faults in BiCMOS devices cause one or more parts (p-part or n-parts) of the circuit to exhibit a different state (conducting or nonconducting) from the fault-free circuit. An input pattern classification scheme is presented for different faults. These classes of patterns are then used to obtain test sets.>

The fiber distributed data interface (FDDI) standard supports synchronous and asynchronous data transmissions. It allows each station to have multiple classes of asynchronous data, and meets the requirements of different classes by means of a timer-based priority scheme. The performance of a FDDI network carrying multiple classes of traffic is considered. An analytical model is presented to evaluate the throughput of synchronous traffic and asynchronous traffic. The model can be used to evaluate the throughput of individual priority classes and the mean token-cycle time of the network, when the network offered load varies from very low values to high values. The governing equations for the throughput characteristics and mean token-cycle time are strictly functions of network parameters. Simulations are used to verify the analytical model. Simulation results are also used to examine the variation of the mean access delay characteristics of various classes and the token-cycle time distribution of the priority scheme. Limitations of the model are outlined.>

Fiber distributed data interface-II (FDDI-II) is a high speed LAN for integrated services applications, supporting isochronous, synchronous, and asynchronous traffic. The performance of FDDI-II in handling multiple classes of traffic is examined. The network can be tuned to meet the requirements of isochronous, synchronous, and asychronous traffic using the wideband channels (WBCs), synchronous allocation, and the target token rotation time. Conditions for providing a guaranteed bandwidth for asynchronous traffic while meeting the requirements of isochronous and synchronous classes are presented. Simulation results are used to evaluate the proposed bandwidth allocation method.

Register insertion bus (RIB) is a high-speed fiber optic network that uses a folded-bus topology and a medium access interface. The protocol is implemented with stations containing two buffers that allow packets in the network to be temporarily stored to avoid conflicts. Access to the network is gained via active taps at each station. Results show the access scheme to efficiently utilize the network without requiring the overhead associated with a train-type protocol. The protocol is scalable in both transmission speed and packet length. An increase in transmission speed results in a proportional increase in throughput. RIB favors upstream stations at high loads. Thus, extensions to the RIB access scheme that use an increased buffer size and information gathered from the network to provide a fair distributed access scheme for all loads are presented. The control is based on increasing or decreasing the time between transmissions, based on a count of network packet transmissions. Results demonstrate that a fair access to the network is achieved for all loads.

The performance of the timer-based priority scheme of the fiber distributed data interface (FDDI) network is investigated. The token-cycle time distribution illustrating the mechanics of the priority scheme is presented. Analytical expressions are given to evaluate the first and the second moments of the token-cycle time for FDDI networks with multiple classes of traffic. An iterative procedure that uses the knowledge of the throughput characteristics is utilized to solve these analytical expressions. The worst case confidance intervals are given. Simulation results are presented for homogeneous and heterogeneous networks with multiple classes of traffic highlighting the token-cycle time variation when the network load changes from a low value to a substantially high value.

BiCMOS devices exhibit sequential behavior under transistor stuck-open (s-open) faults. In addition to the sequential behavior, delay faults are also present. Detection of s-open faults exhibiting sequential behavior need two or multipattern sequences, and delay faults are all the more difficult to detect. A new design for testability scheme is presented for single BJT BiCMOS logic gates which uses only two extra transistors to improve the circuit testability regardless of timing skews/delays, glitches or charge sharing among internal nodes. It requires only a single vector instead of the two or multipattern sequences. The testable design scheme presented also avoids the requirement of generating tests for delay faults.>

It is often assumed that the faults in storage elements (SEs) can be modeled as output/input stuck-at faults of the element. They are implicitly considered equivalent to the stuck-at faults in the combinational logic surrounding the SE cells. Transistor-level faults in common SEs are examined here. A more accurate higher level fault model for elementary SEs that better represents the physical failures is presented. It is shown that a minimal (stuck-at) model may be adequate if only modest fault coverage is desired. The enhanced model includes some common fault behaviors of SEs that are not covered by the minimal fault model. These include data-feedthrough and clock-feedthrough behaviors, as well as problems with logic level retention. Fault models for complex SE cells can be obtained without a significant loss of information about the structure of the circuit. The detectability of feedthrough faults is considered.>

An optimum application-specific data flow architecture for accelerating the standard event driven logic simulation is developed. A conservative distributed simulation algorithm which minimizes the use of NULL messages is also developed, along with a pseudodynamic data flow architecture to implement this distributed algorithm efficiently. A comparison of the standard event driven algorithm-based data flow accelerator to the distributed simulation algorithm-based accelerator is made on several benchmark circuits. The distributed simulation algorithm on the specialized data flow accelerator outperforms the standard event driven algorithm based data flow accelerator by a factor of three in most cases.>

The register insertion bus (RIB) access scheme, a protocol that uses a folded-bus topology such as in C-NET and D-Net but reduces overhead by using an approach similar to that used in the interface of the register insertion ring, is described. The protocol uses stations containing two buffers that allow packets in the network to be temporarily stored to avoid conflicts. Performance results based on analysis and simulation are presented. Results show the access scheme to utilize the network efficiently without requiring the overhead associated with a train-type protocol. The protocol is scalable in both transmission speed and packet length. An increase in transmission speed will result in a proportional increase in performance. This protocol serves as the basis for a protocol that uses information gathered from the network to provide a fair distributed access scheme for all loads. The control is based on incrementally increasing or decreasing the time between transmissions based on the measured network load.>

The logic behavior of single BJT BiCMOS devices under transistor level shorts and opens is examined. In addition to delay faults, faults that cause the gate to exhibit sequential behavior were observed. Several faults can be detected only by monitoring the current. The faulty behaviour of bipolar (TTL) and CMOS logic families is compared with BiCMOS. Effects of bridging faults in BiCMOS devices has been examined for both hard short as well as bridging with a significant resistance.>

Reprogrammable dataflow neural classifiers are proposed as an alternative to traditional implementations. In general, these classifiers are based on functional languages, neural-dataflow transformations, dataflow algorithmic transformations, and dataflow multiprocessors. An experimental approach is used to investigate the performance of a large-scale fine-grained dataflow classifier architecture. In this study, the functional descriptions of high level data dependency of a supervised learning algorithm are transformed into a machine executable low-level dataflow graph. The tagged token dataflow algorithmic transformation is applied to exploit the parallelism. Dataflow neural classifiers are used to implement the learning algorithm. No attempt is made to optimize the granularity of the high-level language programming blocks to balance the computation and communication. The proposed classifier architecture is more versatile than other existing architectures. Performance results show the effectiveness of dataflow neural classifiers.>

Bipolar emitter coupled logic (ECL) devices can now be fabricated at high densities and lower power consumption. With the achievement of low power and high densities, ECL technology is expected to be used widely in high performance digital circuits. This necessitates the need for obtaining optimum gate level models for ECL circuits. A simple technique to obtain a gate level model of an ECL circuit is presented. The gate level models obtained for 1-level and 2-level ECL using the transformation rules presented are the same as the fault models that provide higher coverage of physical failures.>

Current drawn by a static CMOS VLSI integrated circuit during quiescent periods is extremely small and is normally of the order of nanoamperes. However, it is remarkably susceptible to a number of failure modes. Many faults present in such ICs cause the quiescent power-supply current (IDDQ) to increase by several orders of magnitude. Some of these faults may not manifest as logical faults, and would not be detected by traditional IC test techniques. In large ICs, it may be hard to distinguish between larger IDDQ due to defects and elevated IDDQ due to normal parameter variations. A statistical characterisation of the problem is presented. This can be used to determine the optimal size of partitions. A new information compression scheme is presented which can significantly enhance resolution.>

The operation induced faults in CMOS circuits are discussed. The significance of these faults for high density, small geometry circuits is pointed out. For modeling purposes the effects of these faults are correlated with classical fault models. A conductance fault model is presented to incorporate these faults. A test scheme to detect these faults is suggested which is based on the measurement of supply current. A scheme to generate test patterns for these faults is also outlined.>

An analytical model is presented to evaluate the throughput of synchronous traffic and asynchronous traffic. This model can be used to evaluate the throughput of individual priority classes and the mean token-cycle time of the network. A method of selecting timers to support integrated synchronous and asynchronous traffic in FDDI (fiber distributed data interface) is presented. Simulations are used to verify the analytical model. Simulation results are also used to examine the variation of the mean message-delay characteristics of various classes. Limitations of the model are outlined.>

CMOS combinational circuits exhibit sequential behavior in the presence of open faults, thus making it necessary to use two pattern tests. Two or multi-pattern sequences may fail to detect CMOS stuck-open faults in the presence of glitches. The available methods for augmenting CMOS gates to test CMOS stuck-open faults, are found to be inadequate in the presence of glitches. A new CMOS testable design is presented. The scheme uses two additional MOSFETs, which convert a CMOS gate to either pseudo nMos or pseudo pMOS gate during testing. The proposed design ensures the detection of stuck-open faults using a single vector during testing

Switch-level models are widely used for fault analysis of MOS digital circuits. Switch-level analysis (SLA) provides significantly more accurate results compared to gate-level models, and also avoids the complexities of circuit-level analysis. The accuracy of SLA is critically examined, and conditions under which SLA may generate incorrect results are specified. Such conditions may occur when the bulk of a transistor is connected to its source. These conditions are especially applicable under certain types of bridging faults. A simple technique is suggested for accurate switch-level modeling under such conditions.>

The IEEE 802.4 token bus standard defines an optional priority scheme to handle multiple classes of data. It allocates the channel bandwidth among different priority classes of messages by a set of timers at each station. An analytical model for the priority scheme is presented. The model relates the throughput of each priority class of messages to the traffic intensities of different classes, the target rotation times, and the high-priority token holding time. The network is assumed to be symmetric with respect to its parameters and the traffic distribution among nodes. Simulation results are used to evaluate the accuracy of the model. The model provides means of evaluating the network throughput and can be used to determine the time values to meet the throughput requirements of different classes of traffic.>

In the above correspondence (ibid., vol.COM-34, no.12, p.1253-6, Dec. 1986), throughput bounds are derived for the IEEE 802-4 token passing scheme, in the presence of a single class of messages. It does not consider the fact that even after the token holdtimer expires, a node is allowed to complete its message transmission. Here, the fact is considered, and the upper bounds on throughput are modified.>

Bridging faults have been shown to be a major failure mode in VLSI devices. This study examines nMOS and CMOS complex gates in detail for bridging faults. Analysis is carried out using both switch and circuit level models for comparison. It is shown that in most cases, the switch level analysis predicts the correct behavior. A set of conditions are presented, under which the switch level analysis may fail to predict the correct behavior. These conditions can be used for accurate switch level test generation and simulation.