Layered Architecture

There are various ways that network functionalities can be allocated to different layers and to different network elements, some being more desirable than others. The intellectual goal of the research surveyed by this article is to provide a theoretical foundation for these architectural decisions in networking.

We show how nonlinear embedding algorithms popular for use with shallow semisupervised learning techniques such as kernel methods can be applied to deep multilayer architectures, either as a regularizer at the output layer, or on each layer of the architecture. This provides a simple alternative to existing approaches to deep learning whilst yielding competitive error rates compared to those methods, and existing shallow semi-supervised techniques.

location-aware, device-aware and personalized applications. Such applications constantly need to monitor the environment -called context -to allow the application to react accordingly to this context. Context-awareness is especially interesting in mobile scenarios where the context of the application is highly dynamic and allows the application to deal with the constraints of mobile devices in terms of presentation and interaction abilities and communication restrictions.

With increased global interconnectivity, reliance on e-commerce, network services, and Internet communication, computer security has become a necessity. Organizations must protect their systems from intrusion and computer-virus attacks. Such protection must detect anomalous patterns by exploiting known signatures while monitoring normal computer programs and network usage for abnormalities. Current antivirus and network intrusion detection (ID) solutions can become overwhelmed by the burden of capturing and classifying new viral stains and intrusion patterns. To overcome this problem, a self-adaptive distributed agent-based defense immune system based on biological strategies is developed within a hierarchical layered architecture. A prototype interactive system is designed, implemented in Java, and tested. The results validate the use of a distributed-agent biological-system approach toward the computer-security problems of virus elimination and ID.

The paper describes the design, implementation, and evaluation of Depot, a cloud storage system that minimizes trust assumptions. Depot tolerates buggy or malicious behavior by any number of clients or servers, yet it provides safety and liveness guarantees to correct clients. Depot provides these guarantees using a two-layer architecture. First, Depot ensures that the updates observed by correct nodes are consistently ordered under Fork-Join-Causal consistency (FJC). FJC is a slight weakening of causal consistency that can be both safe and live despite faulty nodes. Second, Depot implements protocols that use this consistent ordering of updates to provide other desirable consistency, staleness, durability, and recovery properties. Our evaluation suggests that the costs of these guarantees are modest and that Depot can tolerate faults and maintain good availability, latency, overhead, and staleness even when significant faults occur.

At the MIT Arti cial Intelligence L aboratory we have been working on technologies for an Intelligent Room. Rather than pull people into the virtual world of the computer we are trying to pull the computer out into the real world of people. To do this we are combining robotics and vision technology with speech understanding systems, and agent based a r chitectures to provide ready at hand computation and information services for people engaged in day to day activities, both on their own and in conjunction with others.

Substitutions of sulfate and aluminum ions in calcium silicate hydrate (tobermorite) were investigated at 175 °C under saturated steam pressure. Structure and morphology of different tobermorite samples were investigated using XRD, FTIR and SEM. A sample of pure xonotlite was prepared to assist in FTIR band assignment of different tobermorites. Al3+ decreases the degree of crystallinity at short curing time and increases crystallinity at longer curing times, with a sort of stabilization of small crystals. Al3+ changes the morphology of tobermorite from platy-shape to lath-shape. Sulfate ion increases imperfection of tobermorite crystal structure and the morphology of tobermorite change magnificently to leafy shape. The FTIR spectra of all 1.1 nm tobermorite samples are nearly similar. All tobermorite samples have a significant concentration of Q3 Si sites, which indicates cross-linking of the chains. Sulfate substitution for silicate in tobermorite structure is associated with the appearance of the bands at 1610 and 631 cm−1 which indicates the incorporation of CaO–H group in tobermorite structure. This indicated the charge compensation mechanism; SiO44− is substituted by SO42− and 2OH−.

O ce delivery robots have to perform many tasks. They have to determine the order in which to visit ofces, plan paths to those o ces, follow paths reliably, and avoid static and dynamic obstacles in the process. Reliability and e ciency are key issues in the design of such autonomous robot systems. They must deal reliably with noisy sensors and actuators and with incomplete knowledge of the environment. They must also act e ciently, in real time, to deal with dynamic situations. Our architecture is composed of four abstraction layers: obstacle avoidance, navigation, path planning, and task scheduling. The layers are independent, communicating processes that are always active, processing sensory data and status information to update their decisions and actions. A version of our robot architecture has been in nearly daily use in our building since December 1995. As of July 1996, the robot has traveled more than 75 kilometers in service of over 1800 navigation requests that were speci ed using our World Wide Web interface.

Transforming models is a critical activity in Model Driven Engineering (MDE). With the expected adoption of the OMG QVT standard for model transformation language it is anticipated that the experience in applying model transformations in various cases will increase. However, the QVT standard is just one possible approach to solving model transformation problems. In parallel with the QVT activity many research groups and companies have been working on their own model transformation approaches and languages. It is important for software developers to be able to compare and select the most suitable languages and tools for a particular problem. This paper compares the proposed QVT language and the ATLAS Transformation Language (ATL) as a step in the direction of gathering knowledge about the existing model transformation approaches. The focus is on the major language components (sublanguages and their features, execution tools, etc.) and how they are related. Both languages expose a layered architecture for organizing their components. The paper analyzes the layers and compares them according to various categories. Furthermore, motivations for interoperability between the languages and the related tools are given. Possible solutions for interoperability are identified and discussed.

Wireless sensor networks are being developed for a variety of applications. With the continuing advances in network and application design, appropriate middleware is needed to provide both standardized and portable system abstractions and the capability to support and coordinate concurrent applications on sensor networks. In this paper, we first identify several design principles for such a middleware. These principles motivate a cluster-based lightweight middleware framework that separates application semantics from the underlying hardware, operating system, and network infrastructure. We propose a layered architecture for each cluster that consists of a cluster control layer and a resource management layer. Key design issues and related challenges within this framework that deserve further investigation are outlined. Finally, we discuss a technique for energy-efficient resource allocation in a single-hop cluster, which serves as the basic primitive for the development of the resource management layer.

In this article we explore various network layer concepts that play a crucial role in the design of mobile networking systems. We show that mobility is essentially an address translation problem and is best resolved at the network layer. We describe services that must be supported at the network layer to carry out the task of address translation. Using these service primitives as building blocks, we describe a network-layer architecture which enables smooth integration of mobile end systems within the existing Internet. A summary of some of the key Mobile IP proposals is presented, and it is shown that each proposal can be viewed as a special case of the architecture outlined in this article.

We show that neural network classifiers with single-layer training can be applied efficiently to complex real-world classification problems such as the recognition of handwritten digits. We introduce the STEPNET procedure, which decomposes the problem into simpler subproblems which can be solved by linear separators.

We address the challenge of designing a quantum computer architecture with a layered framework that is modular and facilitates faulttolerance. The framework is flexible and could be used for analysis and comparison of differing quantum computer designs. Using this framework, we develop a complete, layered architecture for quantum computing with optically controlled quantum dots, showing how a myriad of technologies must operate synchronously to achieve fault-tolerance. Our design deliberately takes advantage of the large possibilities for integration afforded by semiconductor fabrication. Quantum information is stored in the electron spin states of a charged quantum dot controlled by ultrafast optical pulses. Optical control makes this system very fast, scalable to large problem sizes, and extensible to quantum communication or distributed architectures. The design of this quantum computer centers on error correction in the form of a topological surface code, which requires only local and nearest-neighbor gates. We analyze several important issues of the surface code that are relevant to an architecture, such as resource accounting and the use of Pauli frames. Furthermore, we investigate the performance of this system and find that Shor's factoring algorithm for a 2048-bit number can be executed in approximately one week.

One of the main challenges already faced by communication networks is the efficient management of increasing complexity. The recently proposed concept of cognitive network appears as a candidate that can address this issue. In this paper, we survey the existing research work on cognitive networks, as well as related and enabling techniques and technologies. We start with identifying the most recent research trends in communication networks and classifying them according to the approach taken towards the traditional layered architecture. In the analysis we focus on two related trends: cross-layer design and cognitive networks. We classify the cognitive networks related work in that mainly concerned with knowledge representation and that predominantly dealing with the cognition loop. We discuss the existing definitions of cognitive networks and, with respect to those, position our understanding of the concept. Next, we provide a summary of artificial intelligence techniques that are potentially suitable for the development of cognitive networks, and map them to the corresponding states of the cognition loop. We summarize and compare seven architectural proposals that comply with the requirements for a cognitive network. We discuss their relative merits and identify some future research challenges before we conclude with an overview of standardization efforts.

This paper describes the development of two buffer layer architectures on rolled-Ni substrates using an electron beam evaporation technique. The first buffer layer architecture consists of an epitaxial laminate of CeO2/Pd/Ni. The second alternative buffer layer consists of an epitaxial laminate of YSZ/CeO2/Ni. The cube (100) texture in the Ni was produced by cold-rolling followed by recrystallization. The CeO 2 films were grown epitaxially on both Pd-buffered and textured-Ni substrates. The YSZ films were grown epitaxially on CeO2-buffered Ni substrates. The crystallographic orientation of the Pd, CeO 2, and YSZ films were all (100). We also studied the effect of the CeO 2 layer thickness and crack formation on textured-Ni substrates. The layer thickness was found to be critical. For some thicknesses, cracks formed in the CeO 2 layer. The presence of YSZ layers on the CeO 2 layers seem to alleviate the cracks that are formed underneath. Our SEM studies showed that both CeO 2 (3-10 nm thick underlayer) and YSZ layers were smooth and continuous.

The provision of medical care typically involves a number of individuals, located ina number of different institutions, whose decisions and actions need to be coordinated if the care is to be effective and efficient. To facilitate this decision making and to ensure the coordination process runs smoothly, the use of software support is becoming increasingly widespread. Tothis end, this

Network protocols in layered architectures have historically been obtained on an ad-hoc basis, and much of the recent cross-layer designs are conducted through piecemeal approaches. Network protocols may instead be holistically analyzed and systematically designed as distributed solutions to some global optimization problems in the form of generalized Network Utility Maximization (NUM), providing insight on what they optimize and on the structures of network protocol stacks. In the form of 10 Questions and Answers, this paper presents a short survey of the recent efforts towards a systematic understanding of "layering" as "optimization decomposition". The overall communication network is modeled by a generalized NUM problem, each layer corresponds to a decomposed subproblem, and the interfaces among layers are quantified as functions of the optimization variables coordinating the subproblems. Furthermore, there are many alternative decompositions, each leading to a different layering architecture. Industry adoption of this unifying framework has also started. Here we summarize the current status of horizontal decomposition into distributed computation and vertical decomposition into functional modules such as congestion control, routing, scheduling, random access, power control, and coding. We also discuss under-explored future research directions in this area. More importantly than proposing any particular crosslayer design, this framework is working towards a mathematical foundation of network architectures and the design process of modularization.

The crystal growth behaviour of solid phase halides encapsulated within single walled carbon nanotubes (SWNTs) is reviewed. As SWNTs form atomically thin channels within a restricted diameter range, their internal van der Waals surfaces regulate the growth behaviour of encapsulated materials in a very precise fashion. Crystal growth within SWNTs is therefore atomically regulated and nano-scale crystals with precise integral layer architectures (i.e. 'Feynman Crystals') can be formed. The structural properties of the resulting 1D crystals are principally dictated by the structural chemistry of the bulk material although deviations from bulk crystal growth behaviour are observed. Alternatively, 1D crystals with completely novel structures can form inside SWNTs. Where the encapsulated crystal has a structure recognizably related to that of the bulk material, crystals are formed with lower surface coordination and all exhibit substantial lattice distortions as a result of this reduced coordination and/or van der Waals constriction effects. 1D crystal growth within SWNTs is occasionally impeded by the presence of simultaneously incorporated fullerene molecules.

The improvement in oxygen barrier properties of poly(ethylene terephthalate) (PET) by incorporation of an impermeable phase such as crystallinity or talc platelets was examined. Crystallinity was induced by crystallization from the glassy state (cold crystallization). Microlayering was used to create talc-filled structures with controlled layer architecture. The reduction of permeability in crystallized and talcfilled PET was well described by Nielsen's model. Changes in permeability of crystalline PET could not be ascribed to the filler effect of crystallites only. Our data on solubility, obtained on the basis of measurements of the oxygen transport coefficients, confirmed a previous finding that the amorphous phase density of PET decreases upon crystallization. The data were amenable to interpretation by free volume theory. Talc-filled materials processed by different methods showed the same permeability; however, much better mechanical properties were achieved by microlayering.

The goal of transfer learning is to use the knowledge acquired in a set of source tasks to improve performance in a related but previously unseen target task. In this paper, we present a multilayered architecture named CAse-Based Reinforcement Learner (CARL). It uses a novel combination of Case-Based Reasoning (CBR) and Reinforcement Learning (RL) to achieve transfer while playing against the Game AI across a variety of scenarios in MadRTS TM , a commercial Real Time Strategy game. Our experiments demonstrate that CARL not only performs well on individual tasks but also exhibits significant performance gains when allowed to transfer knowledge from previous tasks.

Building automation (BA) and smart homes (SHs) have traditionally not been a unified field but varied by their origins, legal foundations, different applications, different goals, and national funding programs for basic research. Only within the last years that an international common focus appeared. The following overview gives not only an introduction into the topic of BA but also the distinction to other areas of automation, in which networks of the field level (the sensor and actuator level) play an important role. Finally, the scientific challenges will be mentioned. SHs are referred to when the differences to BA have to be explicitly stressed. This paper is an introduction for the special IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS section on BA and shall introduce the reader to this new topic. BA not only has a huge economic potential but also is of significant academic interest today.

We present HyperGraphDB, a novel graph database based on generalized hypergraphs where hyperedges can contain other hyperedges. This generalization automatically reifies every entity expressed in the database thus removing many of the usual difficulties in dealing with higher-order relationships. An open two-layered architecture of the data organization yields a highly customizable system where specific domain representations can be optimized while remaining within a uniform conceptual framework. HyperGraphDB is an embedded, transactional database designed as a universal data model for highly complex, large scale knowledge representation applications such as found in artificial intelligence, bioinformatics and natural language processing.

We address the semantic gap problem in behavioral monitoring by using hierarchical behavior graphs to infer high-level behaviors from myriad low-level events. Our experimental system traces the execution of a process, performing data-flow analysis to identify meaningful actions such as "proxying", "keystroke logging", "data leaking", and "downloading and executing a program" from complex combinations of rudimentary system calls. To preemptively address evasive malware behavior, our specifications are carefully crafted to detect alternative sequences of events that achieve the same high-level goal. We tested eleven benign programs, variants from seven malicious bot families, four trojans, and three mass-mailing worms and found that we were able to thoroughly identify high-level behaviors across this diverse code base. Moreover, we effectively distinguished malicious execution of high-level behaviors from benign by identifying remotely-initiated actions.

In this article, we will present an overview of the Coupled Layered Architecture for Robotic Autonomy. CLARAty develops a framework for generic and reusable robotic components that can be adapted to a number of heterogeneous robot platforms. It also provides a hmework that will simplify the integration of new technologies and enable the comparison of various elements. CLARAty consists of two distinct layers: a FUIIC~~OM~ Layer and a Decision Layer. The Functional Layer defines the various abstractions of the system and adapts the abstract components to real or simulated devices. It provides a-work and the algorithms for low-and mid-level autonomy. The Decision Layer provides the system's high-level autonomy, which reasons about global resources and mission constraints. The Decision Layer accesses information from the Functional Layer at multiple levels of granularity. In this article, we will also present some of the challenges in developing interoperable software for various rover platforms. Examples will include challenges from the locomotion and manipulation domains

We report high efficiency and low roll-off for blue electrophosphorescent organic light emitting devices based on a mixed host layer architecture. The devices were fabricated using a mixed layer of di-͓4-͑N,N-ditolyl-amino͒-phenyl͔cyclohexane, a hole transport material, and 2,8-bis͑diphenylphosphoryl͒dibenzothiophene, an electron transport material, as the host layer doped with the blue phosphor iridium ͑III͒ bis͓͑4,6-difluorophenyl͒-pyridinato-N , C 2Ј ͔picolinate. Using a mixed layer as the host allowed us to achieve high power efficiency ͑59 lm/W at 100 cd/ m 2 ͒, low turn-on voltage ͑2.7 V for Ͼ10 cd/ m 2 ͒, and low roll-off in these devices.

Rules and ontologies play a key role in the layered architecture of the Semantic Web, as they are used to ascribe meaning to, and to reason about, data on the Web. While the Ontology Layer of the Semantic Web is quite developed, and the Web Ontology Language (OWL) is a W3C recommendation since a couple of years already, the rules layer is far less developed and an active area of research; a number of initiatives and proposals have been made so far, but no standard as been released yet. Many implementations of rule engines are around which deal with Semantic Web data in one or another way. This article gives a comprehensive, although not exhaustive, overview of such systems, describes their supported languages, and sets them in relation with theoretical approaches for combining rules and ontologies as foreseen in the Semantic Web architecture. In the course of this, we identify desired properties and common features of rule languages and evaluate existing systems against their support. Furthermore, we review technical problems underlying the integration of rules and ontologies, and classify representative proposals for theoretical integration approaches into different categories.

The purpose of this paper is to study how Intelligent Agents IAs can be used to facilitate electronic trading. An IA is a software program designed for performing a specific task based on its own knowledge and the message it received. Given the increased complexity of Internet services, many IAs are useful to make electronic markets more effective.

We present in this paper a new set of metrics that measure the quality of modularization of a non-object-oriented software system. We have proposed a set of design principles to capture the notion of modularity and defined metrics centered around these principles. These metrics characterize the software from a variety of perspectives: structural, architectural, and notions such as the similarity of purpose and commonality of goals. (By structural, we are referring to intermodule coupling-based notions, and by architectural, we mean the horizontal layering of modules in large software systems.) We employ the notion of API (Application Programming Interface) as the basis for our structural metrics. The rest of the metrics we present are in support of those that are based on API. Some of the important support metrics include those that characterize each module on the basis of the similarity of purpose of the services offered by the module. These metrics are based on information-theoretic principles. We tested our metrics on some popular open-source systems and some large legacy-code business applications. To validate the metrics, we compared the results obtained on human-modularized versions of the software (as created by the developers of the software) with those obtained on randomized versions of the code. For randomized versions, the assignment of the individual functions to modules was randomized.

The paper presents the framework for agent-enabled dynamic web service composition. The core of the methodology is the new understanding of a web service as an agent capability having proper ontological description. It is demonstrated how diverse web services may be composed and mediated by dynamic coalitions of software agents collaboratively performing tasks for service requestors. Middle Agent Layer is introduced to conduct service request to task transformation, agent-enabled cooperative task decomposition and performance. Discussed are the formal means to arrange agents' negotiation, to represent the semantic structure of the task-activity-service hierarchy and to assess fellow-agents' capabilities and credibility factors. Finally, it is argued that the presented formal technique is applicable to various application domains. Presented is the ongoing work on designing and implementing agent-based layered architecture for intelligent rational information and document retrieval. Finally, the discussion of the OntoServ.Net framework for the development of P2P mobile service infrastructures for industrial asset management provides the extension of the web service composition approach.

A number of analogies to cryptographic concepts have been made about watermarking. In this paper, we argue that these analogies are misleading or incorrect, and highlight several analogies to support our argument. We believe that the fundamental role of watermarking is the reliable embedding and detection of information and should therefore be considered a form of communications. We note that the fields of communications and cryptography are quite distinct and while communications systems often combine technologies from the two fields, a layered architecture is applied that requires no knowledge of the layers above. We discuss how this layered approach can be applied to watermarking applications.

In this study, normal EEG series recorded from healthy volunteers and epileptic EEG series recorded from patients within and without seizure are classified by using Multilayer Neural Network (MLNN) architectures with respect to several time domain entropy measures such as Shannon Entropy (ShanEn), Log Energy Entropy (LogEn), and Sample Entropy (Sampen). In tests, the MLNN is performed with several numbers of neurons for both one hidden layer and two hidden layers. The results show that segments in seizure have significantly lower entropy values than normal EEG series. This result indicates an important increase of EEG regularity in epilepsy patients. The LogEn approach, which has not been experienced in EEG classification yet, provides the most reliable features into the EEG classification with very low absolute error as 0.01. In particular, the MLNN can be proposed to distinguish the seizure activity from the seizure-free epileptic series where the LogEn values are considered as signal features that characterize the degree of EEG complexity. The highest classification accuracy is obtained for one hidden layer architecture.

The Internet of Things (IoT) is leading today's digital transformation. Relying on a combination of technologies, protocols, and devices such as wireless sensors and newly developed wearable and implanted sensors, IoT is changing every aspect of daily life, especially recent applications in digital healthcare. IoT incorporates various kinds of hardware, communication protocols, and services. This IoT diversity can be viewed as a double-edged sword that provides comfort to users but can lead also to a large number of security threats and attacks. In this survey paper, a new compacted and optimized architecture for IoT is proposed based on five layers. Likewise, we propose a new classification of security threats and attacks based on new IoT architecture. The IoT architecture involves a physical perception layer, a network and protocol layer, a transport layer, an application layer, and a data and cloud services layer. First, the physical sensing layer incorporates the basic hardware used by IoT. Second, we highlight the various network and protocol technologies employed by IoT, and review the security threats and solutions. Transport protocols are exhibited and the security threats against them are discussed while providing common solutions. Then, the application layer involves application protocols and lightweight encryption algorithms for IoT. Finally, in the data and cloud services layer, the main important security features of IoT cloud platforms are addressed, involving confidentiality, integrity, authorization, authentication, and encryption protocols. The paper is concluded by presenting the open research issues and future directions towards securing IoT, including the lack of standardized lightweight encryption algorithms, the use of machine-learning algorithms to enhance security and the related challenges, the use of Blockchain to address security challenges in IoT, and the implications of IoT deployment in 5G and beyond.

In this article, we describe a multi-layered architecture for sketch-based interaction within virtual environments. Our architecture consists of eight hierarchically arranged layers that are described by giving examples of how they are implemented and how they interact. Focusing on table-like projection systems (such as Virtual Tables or Responsive Workbenches) as human-centered output-devices, we show examples of how to integrate parts or all of the architecture into existing domain-speci"c applications * rather than realizing new general sketch applications * to make sketching an integral part of the next-generation human}computer interface. : S 0 0 9 7 -8 4 9 3 ( 0 0 ) 0 0 0 8 8 -1

Data-centric scientific workflows are often modeled as dataflow process networks. The simplicity of the dataflow framework facilitates workflow design, analysis, and optimization. However, modeling "control-flow intensive" tasks using dataflow constructs often leads to overly complicated workflows that are hard to comprehend, reuse, and maintain. We describe a generic framework, based on scientific workflow templates and frames, for embedding control-flow intensive subtasks within dataflow process networks. This approach can seamlessly handle complex control-flow without sacrificing the benefits of dataflow. We illustrate our approach with a real-world scientific workflow from the astrophysics domain, requiring remote execution and file transfer in a semi-reliable environment. For such workflows, we also describe a 3-layered architecture based on frames and templates where the top-layer consists of an overall dataflow process network, the second layer consists of a tranducer template for modeling the desired control-flow behavior, and the bottom layer consists of frames inside the template that are specialized by embedding the desired component implementation. Our approach can enable scientific workflows that are more robust (faulttolerance strategies can be defined by control-flow driven transducer templates) and at the same time more reusable, since the embedding of frames and templates yields more structured and modular workflow designs.

This paper presents an architecture that enables multiple robots to explicitly coordinate actions at multiple levels of abstraction. In particular, we are developing an extension to the traditional three-layered robot architecture that enables robots to interact directly at each layer-at the behavioral level, the robots create distributed control loops; at the executive level, they synchronize task execution; at the planning level, they use market-based techniques to assign tasks, form teams, and allocate resources. We illustrate these ideas through applications in multi-robot assembly, multi-robot deployment, and multi-robot mapping.

Process management is a method for improving Boeing's business processes, however many aspects have been difficult to implement. eQuality is a software system based on a framework called DDUCKS that is being designed to support the process management life cycle. We take a knowledge acquisition approach to the development of the tool, emphasizing the importance of mediating and intermediate knowledge representations. Sharing and reuse of tools, models, and representations is facilitated through a layered architecture. eQuality's process documentation capability includes a number of views, that can be used either in sketchpad or model mode. Using the views, an integrated business enterprise model may be developed. Analysis and simulation tools supporting process improvement are implemented with attribute, function, and task editors that make use of a user scripting language and extensible function library. A virtual project notebook is used to organize project information and help facilitate group meetings.

In this paper, we present an open-source platform for wireless body sensor networks called DexterNet. The system is motivated by shifting research paradigms to support real-time, persistent human monitoring in both indoor and outdoor environments. The platform utilizes a three-layer architecture to control heterogeneous body sensors. The first layer, called the body sensor layer (BSL), deals with design of different wireless body sensors and their instrumentation on the body. We detail two custom-built body sensors: one measuring body motions and the other measuring the ECG and respiratory patterns. At the second layer, called the personal network layer (PNL), the wireless body sensors on a single subject communicate with a mobile base station, which supports Linux OS and the IEEE 802.15.4 protocol. The BSL and PNL functions are abstracted and implemented as an opensource software library, called Signal Processing In Node Environment (SPINE). A DexterNet network is scalable, and can be reconfigured on-the-fly via SPINE. At the third layer, called the global network layer (GNL), multiple PNLs communicate with a remote Internet server to permanently log the sensor data and support higher-level applications. We demonstrate the versatility of the DexterNet platform via three applications: avatar visualization, human activity recognition, and integration of DexterNet with global positioning sensors and air pollution sensors for asthma studies.

The Internet's layered architecture and organizational structure give rise to a number of different topologies, with the lower layers defining more physical and the higher layers more virtual/logical types of connectivity structures. These structures are very different, and successful Internet topology modeling requires annotating the nodes and edges of the corresponding graphs with information that reflects their network-intrinsic meaning. These structures also give rise to different representations of the traffic that traverses the heterogeneous Internet, and a traffic matrix is a compact and succinct description of the traffic exchanges between the nodes in a given connectivity structure. In this paper, we summarize recent advances in Internet research related to (i) inferring and modeling the router-level topologies of individual service providers (i.e., the physical connectivity structure of an ISP, where nodes are routers/switches and links represent physical connections), (ii) estimating the intra-AS traffic matrix when the AS's router-level topology and routing configuration are known, (iii) inferring and modeling the Internet's AS-level topology, and (iv) estimating the inter-AS traffic matrix. We will also discuss recent work on Internet connectivity structures that arise at the higher layers in the IP/TCP protocol stack and are more virtual and dynamic; e.g., overlay networks like the WWW graph, where nodes are web pages and edges represent existing hyperlinks, or P2P networks like Gnutella, where nodes represent peers and two peers are connected if they have an active network connection. Here, an autonomous system (AS) or autonomous domain is a group of routers and networks managed by a single organization. In turn, an Internet Service Provider (ISP) can consist of a single AS or of a group of ASes, but for simplicity, we will use the terms AS and ISP indistinguishably throughout this paper. The TCP/IP protocol stack as a whole and IP in particular are able to hide from the user much of the enormous complexity associated with controlling this diverse set of networked resources and coordinating the actions among the many competing ISPs. By providing the mechanisms necessary to knit together diverse networking technologies and ASes into a single virtual network (i.e., a network of networks, or "Internet"), they ultimately guarantee seamless connectivity and reliable communication between sending and receiving hosts, irrespective of where in the network they are.

This paper presents a new neural network to solve the shortest path problem for internetwork routing. The proposed solution extends the traditional single-layer recurrent Hopfield architecture introducing a two-layer architecture that automatically guarantees an entire set of constraints held by any valid solution to the shortest path problem. This new method addresses some of the limitations of previous solutions, in particular the lack of reliability in what concerns succeeded and valid convergence. Experimental results show that a clear improvement in well-succeeded convergence can be achieved. Additionally, computation performance is also improved at the expense of slightly worse results.

Optical coherence tomography (OCT) is an optical imaging modality that performs high-resolution, cross- sectional, subsurface tomographic imaging of the microstructure of tissues. The physical principle of OCT is similar to that of B-mode ultrasound imaging, except that it uses infrared light waves rather than acoustic waves. The in vivo resolution is 10-25 times better (about 10 µm) than with high-frequency