Intermediate Network Research Articles

Variable data rate (VDR) network congestion control (NCC) applied to voice/audio communication

This paper presents three novel methods for implementing variable data rate (VDR) network congestion control (NCC), and compares them with traditional random (RND) drop methods. The proposed methods include: frequency (FRQ), combined random-innerloop frequency-outerloop (CRFRQ), and combined frequency-innerloop random-outerloop (CFRND). A generic model is used for RND drop that approximates congestion avoidance methods that drop packets. These methods allow NCC at intermediate network nodes using novel methods for setting data priority values. This is a first step towards demonstrating the need for stronger and efficient router technology to enhance Quality of Service (QoS) for voice applications. In the examples considered in this paper, the FRQ method reduced outage from 60% with RND to 0% with FRQ with 5 incoming flows through a bandwidth limited threshold of 1000 bits. Mean Opinion Score (MOS) was increased by a factor in the range from 0.2 to 1 on the MOS scale when using FRQ and CRFRQ over RND in the tested scenarios. CFRND performs in a similar manner to RND except it allows packet shrinking to increase network utilization. Network utilization increased by up to 37.5% over the RND method compared with the novel methods in one example where all possible outcomes for RND were analyzed. Computational complexity with the novel methods can yield a speedup from 0.02ms faster for 1 flow to 0.56ms faster for 10 flows over traditional lossy compression methods that rely on FFT operations, which are commonly found in lossy voice coders such as Speex, MELPe, and others. Finally, live real-time transport protocol (RTP) congestion tests were performed which confirmed the simulation results that the network congestion control methods achieve higher MOS scores and better QoS.

Heat pump integration in a cheese factory

The aim of process integration is to increase the efficiency and to reduce the energy consumption, operating costs and CO2 emissions of an industrial process. The proposed methodology is applied to a real case study of a cheese factory with non-simultaneous process operations and uses the time average approach combined with restricted matches. This work focuses on appropriate heat pump integration and two different integration strategies are proposed. In the first option, process modifications and direct heat exchange between the process and heat pump streams are not allowed, which means that the integration of heat pumps has to be realized through intermediate heat transfer networks. On the contrary, in the second option, direct heat exchange and process modifications are possible. The results of both options are analyzed and compared for the French and German context. Depending on the industrial constraints and the location, it is shown that the final choice of new heat pump installations may be different. Saving potential in operating costs is higher for option 2 where the cost savings can be higher than 40% for both countries. Furthermore the potential CO2 emissions and primary energy savings are compared.

Video Aggregation: Delivering Videos over Wireless Smart Camera Networks

The proliferation of wireless technologies and inexpensive network-cameras has enabled low-cost and quick deployment of cameras for several surveillance applications, such as traffic monitoring and border control. Smart Camera Networks (SCNs) are networks of cameras that self configure and adapt to improve their operation and reduce the demand on human operators. However, SCNs are constrained by the ability of the underlying wireless network. Streaming video over a network requires substantial bandwidth, and strict Quality-of-Service (QoS) guarantees. In contrast, existing wireless networks have limited bandwidth, and the protocols do not guarantee QoS. Thus, for SCNs to scale beyond a small number of cameras, it is vital to design efficient video delivery protocols that are aware of the limitations of the underlying wireless network. We propose to use Video Aggregation, a technique that enables efficient delivery of video in SCNs by combining related video streams. Existing SCNs use traditional routing protocols where intermediate network routers simply forward the video packets from cameras towards the video analysis center (or base-station). This is inefficient in SCNs since multiple cameras often cover overlapping regions of interest, and video information of these regions are redundantly transmitted over the network. The proposed video aggregation protocol eliminates redundant transmissions by dynamically pruning the overlapping areas at the intermediate routers. The routers blend the received streams into one panoramic video stream with no overlaps. Aggregation also dynamically controls the streaming rate to avoid network congestion and packet drops; the routers adjust the rate of the outgoing video by estimating the available network bandwidth. Thus, base-stations receive video frames with minimal packet drops, thus improving the quality of received video. Our testbed and simulation results show that aggregation outperforms traditional routing both in terms of received video quality and network bandwidth usage. Our testbed experiments show that aggregation improves the received video quality (the Peak Signal-to-Noise-Ratio metric) by 54%. In larger networks, we observed that aggregation eliminates up to 90% of packet drops that were observed in SCNs with traditional routing. In future, we plan to develop a suite of video delivery protocols, which include SCN-aware scheduling and transport protocols.

How attention and reinforcers jointly optimize the associations between sensory representations, working memory and motor programs

Almost all animal behaviors can be seen as sequences of actions towards achieving certain goals. How the association cortices learn to link sensory stimuli to a correct sequence of motor responses is not well understood, especially when only a correct sequence of responses is rewarding. We present a biologically plausible neuronal network model that can be trained to perform a large variety of tasks when only stimuli and reward contingencies are varied. The model’s aim is to learn action values in a feedforward neuronal network and we present mechanisms to overcome the structural and temporal credit assignment problems. The temporal credit assignment problem is solved by a form of Q-learning [1]. The structural credit assignment problem is solved by a form of ‘attentional’ feedback from motor cortex to association cortex that delineates the units that should change connectivity to improve behavior [2]. Moreover, the model has a new mechanism to store traces of relevant sensory stimuli in working memory. During learning, the sensory stimuli, in combination with traces of previous stimuli in working memory become associated with a unique set of action values. Learning in the model is biologically realistic as model units have Hebbian plasticity that is gated by two factors [2]. Firstly, reinforcers or increases in reward expectancy cause the global release of neuromodulatory signals that inform all synapses of the network if the outcome of a trial was better or worse than expected [3]. Selective attention is the second factor that gates plasticity. Attentional feedback highlights the chain of neurons between sensory and motor cortex responsible for the selected action. Only neurons that are causally linked to the action receive attentional feedback, and change the strength of their connections. Selective attention thereby solves the structural credit assignment problem. The resulting learning rule is a form of AGREL [2], which was previously shown to be on average equivalent to error-backpropagation in classification tasks with immediate reward. The present generalization of the learning scheme is based on temporal difference learning and it can train multilayer feedforward networks to perform delayed reward tasks with multiple epochs that require multiple behavioral responses. Importantly, the generalization MQ-AGREL learns to store in working memory information that is relevant at a later stage during a task. This memory is maintained by persistent activity of units at the intermediate network layers. We show that MQ-AGREL can be trained in many tasks that are in use in neurophysiology, including (1) (delayed) saccade-antisaccade tasks; (2) categorization tasks; and (3) probabilistic classification tasks. Neurons at intermediate levels of the network acquire visual responses and memory responses as the result of training that resemble the tuning of neurons in association areas of the cerebral cortex of animals that are trained in these same tasks. We conclude that MQ-AGREL is a powerful and biologically realistic learning rule that accounts for learning in delayed reward tasks that involve non-linear mappings from sensory stimuli and working memory onto motor responses.

Using network properties to predict disease dynamics on human contact networks

Recent studies have increasingly turned to graph theory to model more realistic contact structures that characterize disease spread. Because of the computational demands of these methods, many researchers have sought to use measures of network structure to modify analytically tractable differential equation models. Several of these studies have focused on the degree distribution of the contact network as the basis for their modifications. We show that although degree distribution is sufficient to predict disease behaviour on very sparse or very dense human contact networks, for intermediate density networks we must include information on clustering and path length to accurately predict disease behaviour. Using these three metrics, we were able to explain more than 98 per cent of the variation in endemic disease levels in our stochastic simulations.

Modeling the evolution of a classic genetic switch

BackgroundThe regulatory network underlying the yeast galactose-use pathway has emerged as a model system for the study of regulatory network evolution. Evidence has recently been provided for adaptive evolution in this network following a whole genome duplication event. An ancestral gene encoding a bi-functional galactokinase and co-inducer protein molecule has become subfunctionalized as paralogous genes (GAL1 and GAL3) in Saccharomyces cerevisiae, with most fitness gains being attributable to changes in cis-regulatory elements. However, the quantitative functional implications of the evolutionary changes in this regulatory network remain unexplored.ResultsWe develop a modeling framework to examine the evolution of the GAL regulatory network. This enables us to translate molecular changes in the regulatory network to changes in quantitative network function. We computationally reconstruct an inferred ancestral version of the network and trace the evolutionary paths in the lineage leading to S. cerevisiae. We explore the evolutionary landscape of possible regulatory networks and find that the operation of intermediate networks leading to S. cerevisiae differs substantially depending on the order in which evolutionary changes accumulate; in particular, we systematically explore evolutionary paths and find that some network features cannot be optimized simultaneously.ConclusionsWe find that a computational modeling approach can be used to analyze the evolution of a well-studied regulatory network. Our results are consistent with several experimental studies of the evolutionary of the GAL regulatory network, including increased fitness in Saccharomyces due to duplication and adaptive regulatory divergence. The conceptual and computational tools that we have developed may be applicable in further studies of regulatory network evolution.

On the Capacity of Noncoherent Network Coding

We consider the problem of multicasting information from a source to a set of receivers over a network where intermediate network nodes perform randomized linear network coding operations on the source packets. We propose a channel model for the noncoherent network coding introduced by Koetter and Kschischang in , that captures the essence of such a network operation, and calculate the capacity as a function of network parameters. We prove that use of subspace coding is optimal, and show that, in some cases, the capacity-achieving distribution uses subspaces of several dimensions, where the employed dimensions depend on the packet length. This model and the results also allow us to give guidelines on when subspace coding is beneficial for the proposed model and by how much, in comparison to a coding vector approach, from a capacity viewpoint. We extend our results to the case of multiple source multicast that creates a virtual multiple access channel.

Sub-10 nm structures on silicon by thermal dewetting of platinum

A study of the dewetting behavior of platinum-thin-films on silicon was carried out todetermine how variation of dewetting parameters affects the evolution of film morphologyand to pinpoint which parameters yielded the smallest, most circular features. Platinumfilm thickness as well as dewetting time and temperature were varied and the filmmorphology characterized by means of scanning electron microscopy (SEM) analysis. Twodifferent pathways of dewetting predicted in the literature (Vrij 1966 Discuss. Faraday Soc.42 23, Becker et al 2003 Nat. Mater. 2 59–63) were observed. Depending on the initialcriteria, restructuring of the film occurred via hole or droplet formation. Withincreased annealing time, a transition from an intermediate network structure toseparated islands occurred. In addition, the formation of multilayered films, silicidecrystals and nanowires occurred for certain parameters. Nevertheless, the dewettingbehavior witnessed could be related to physical processes. Droplets with a meandiameter of 9 nm were formed by using a 1.5 nm thick platinum film annealed at800 °C for 30 s. To demonstrate the suitability of the annealed films for further processing, we thenused the dewetted films as masks for reactive ion etching to transfer the pattern into thesilicon substrate, forming tapered nanopillars.

Network Modulation: An Algebraic Approach to Enhancing Network Data Persistence

Large-scale distributed systems such as sensor networks usually experience dynamic topology changes, data losses, and node failures in various catastrophic or emergent environments. As such, maintaining data persistence in a scalable fashion has become critical and essential for such systems. The existing major efforts such as coding, routing, and traditional all have their own limitations. In this work, we propose a novel network (NeMo) approach to significantly improve the data persistence. Built on algebraic number theory, NeMo operates at the level of modulated symbols (so-called modulation over modulation). Its core notion is to mix data at intermediate network nodes and meanwhile guarantee the symbol recovery at the sink(s) without prestoring or waiting for other symbols. In contrast to the traditional thought that n linearly independent equations are needed to solve for n unknowns, NeMo opens a new regime to boost the convergence speed of achieving persistence. Different performance criteria (e.g., and demodulation complexity, convergence speed, finite-bit representation, and noise robustness) have been evaluated in the comprehensive simulations and real experiments to show that the proposed approach is efficient to enhance the network data persistence.

Integrated network experimentation for QoS measurements in opaque MANETs

AbstractIntegrated network experimentation often combines real nodes with simulated ones, each modeling different portions of the topology, in the same experimental run. They enable new validation techniques and larger experiments than obtainable using real elements alone. Integrated experiments can be particularly useful in testing and validating QoS mechanisms for mobile ad hoc networks (MANETs), when the characteristics of the intermediate wireless network segments are not observable from the end segments, and as a consequence these intermediate segments have to be treated as opaque networks; and also when there is a need to conduct experiments in a real MANET environment at a scale larger than just a few nodes. End‐to‐end QoS assurance for such opaque networks, consisting of admission control and quality adjustment, can be based on techniques for dynamically measuring throughput representing the state of these networks. In this paper, we describe a distributed and hybrid testbed that has been deployed for running large‐scale integrated experiments to demonstrate the efficacy of a measurement‐based QoS solution. The infrastructure for the testbed provides an integrated platform consisting of real nodes running the actual software under test, augmented with a simulated network environment. We define a set of metrics and run experiments to evaluate the effectiveness of the QoS solution as well the performance of the deployed testbed. We propose an alternative architecture that employs a Xen‐based virtualization of the real nodes from the deployed testbed. We compare the performances of the virtualized architecture with the deployed architecture vis‐à‐vis latency and resource utilization. Our goal is to establish benchmarks for running large‐scale experiments on performance and QoS measurements in virtualized environments. Copyright © 2009 John Wiley & Sons, Ltd.

The Legality of Deep Packet Inspection

Deep packet inspection is a technology which enables the examination of the content of information packets being sent over the Internet. The Internet was originally set up using “end-to-end connectivity” as part of its design, allowing nodes of the network to send packets to all other nodes of the network, without requiring intermediate network elements to maintain status information about the transmission. In this way, the Internet was created as a “dumb” network, with “intelligent” devices (such as personal computers) at the end or “last mile” of the network. The dumb network does not interfere with an application's operation, nor is it sensitive to the needs of an application, and as such it treats all information sent over it as (more or less) equal. Yet, deep packet inspection allows the examination of packets at places on the network which are not endpoints, In practice, this permits entities such as Internet service providers (ISPs) or governments to observe the content of the information being sent, and perhaps even manipulate it. Indeed, the existence and implementation of deep packet inspection may challenge profoundly the egalitarian and open character of the Internet.This paper will firstly elaborate on what deep packet inspection is and how it works from a technological perspective, before going on to examine how it is being used in practice by governments and corporations. Legal problems have already been created by the use of deep packet inspection, which involve fundamental rights (especially of Internet users), such as freedom of expression and privacy, as well as more economic concerns, such as competition and copyright. These issues will be considered, and an assessment of the conformity of the use of deep packet inspection with law will be made. There will be a concentration on the use of deep packet inspection in European and North American jurisdictions, where it has already provoked debate, particularly in the context of discussions on net neutrality. This paper will also incorporate a more fundamental assessment of the values that are desirable for the Internet to respect and exhibit (such as openness, equality and neutrality), before concluding with the formulation of a legal and regulatory response to the use of this technology, in accordance with these values.

Design and development of framework for diagnosing intermediate nodes

A framework is an integrated system that sets the rules of Automation of a specific product. This system integrates the function libraries, test data sources, object details and various reusable modules. This paper proposes a framework, used for diagnosing and performance analysis of intermediate network nodes such as load balancer, routers, servers etc. For analyzing the performance 'm' number of servers and 'n' number of clients are considered. This framework will help developers working on network nodes to check for the performance of network node component and also to detect the errors in the algorithms.

Inexpensive high availability solutions for the SIP-based VoIP service

In the era of IP-based service, people expect a simple, cheap, and competent Voice over IP (VoIP) service as an alternative of the traditional voice over PSTN. The introduction of the SIP protocol realizes the expectation. Following the cost saving spirit of VoIP, we focus on studying inexpensive high availability solutions for the SIP-based VoIP Service. In this paper, Peer-to-Peer (P2P) based and DN-LB based schemes are mainly compared in the paper. A P2P-based scheme enables an inexpensive high availability solution to the VoIP service by the shared computation resources form P2P nodes. Such a P2P-based solution may be appropriate for an individual VoIP user. However, a caller may take a large volume of messages to find out a callee via the proxy nodes in the P2P network. This inherent property of a P2P network may induce the message overhead and long call setup delay. Based on above, another inexpensive scheme, which is a probing-based name resolution solution, is proposed to achieve high availability and load balancing for the VoIP service. We tag the probing mechanism onto the open source project Domain Name Relay Daemon (DNRD) to become a domain name resolution based load balancer (DN-LB). With DN-LB, all request messages from clients can be fairly distributed to all failure-proof proxy servers in the server farm without using any additional costly intermediate network device and changing the standard SIP architecture. Such a DN-LB based solution may be a good choice for a VoIP service provider.

A method to crystallize substances that oil out

A new approach to crystallize oily substances is described. The tendency for liquid–liquid phase separation (LLPS) is reduced by decreasing the kinetics of self-association via the formation of an intermediate amorphous network. The path to initial crystal formation followed a sequence of first freeze-drying an emulsion of solute in the solvent system followed by suspending the dried solid in water to obtain a hydrated crystalline form. This new procedure was applied successfully to a pharmaceutical organic substance that was previously isolated only as a viscous oil. Once isolated, crystals of the drug were utilized as seeds to allow the successful transformation of an emulsion of the substance into a suspension of crystalline drug solid thus avoiding the freeze-drying step. The isolated crystalline solid retained its physical and chemical purity at room temperature for at least 3 months.

An Extended Review of Techniques for Enhancing TCP Performance

Transmission Control Protocol (TCP) is considered one of the most important protocols in the Internet. An important mechanism in TCP is the congestion control mechanism which controls TCP sending rate and makes TCP react to congestion signals. Nowadays in heteroge- neous networks, TCP may work in networks with some links that have lossy nature (wireless networks for example). TCP treats all packet loss as if they were due to congestion. Consequently, when used in networks that have lossy links, TCP reduces sending rate aggressively when there are transmission (non-congestion) errors in an uncongested network. In this paper we present different solutions to overcome the performance degradation problem TCP faces when working over lossy links. Many solutions have been proposed but we will concentrate on end-to-end so- lutions that require no help from the intermediate network.

Five-level Z-source diode-clamped inverter

This study proposes a five-level Z-source diode-clamped inverter designed with two intermediate Z-source networks connected between the dc input sources and rear-end inverter circuitry. By partially shorting the Z-source networks, new operating states not previously reported for two-level Z-source inverter are introduced here for operating the proposed inverter with voltage buck-boost energy conversion ability and five-level phase voltage switching. These characteristic features are in fact always ensured at the inverter terminal output by simply adopting a properly designed carrier modulation scheme, which always inserts two partial shoot-through states per half carrier cycle for smooth balanced operation. Theoretical findings and practical issues identified are eventually verified by constructing a scaled down laboratory prototype for testing.

SVM-based loss differentiation algorithm for wired-cum-wireless networks

In a hybrid wired-cum-wireless network environment, packet loss may happen because of congestion or wireless link errors. Therefore, differentiating the cause is important for helping transport protocols take actions to control congestion only when the loss is caused by congestion. In this article, an end-to-end loss differentiation mechanism is proposed to improve the transmission performance of transmission control protocol (TCP)-friendly rate control (TFRC) protocol. Its key design is the introduction of the outstanding machine learning algorithm – the support vector machine (SVM) into the network domain to perform multi-metric joint loss differentiation. The SVM is characterized by using end-to-end indicators for input, such as the relative one-way trip time and the inter-arrival time of packets fore-and-aft the loss, while requiring no support from intermediate network apparatus. Simulations are carried out to evaluate the loss differentiation algorithm with various network configurations, such as with different competing flows, wireless loss rate and queue size. The results show that the proposed classifier is effective under most scenarios, and that its performance is superior to the ZigZag, mBiaz and spike (ZBS) scheme.

Intermediate Phases, structural variance and network demixing in chalcogenides: The unusual case of group V sulfides

We review intermediate phases (IPs) in chalcogenide glasses and provide a structural interpretation of these phases. In binary group IV selenides, IPs reside in the 2.40 < r < 2.54 range, and in binary group V selenides they shift to a lower r, in the 2.29 < r < 2.40 range. Here, r represents the mean coordination number of glasses. In ternary alloys containing equal proportions of group IV and V selenides, IPs are wider and encompass ranges of respective binary glasses. These data suggest that the local structural variance contributing to IP widths largely derives from four isostatic local structures of varying connectivity r, two include group V based quasi-tetrahedral ( r = 2.29) and pyramidal ( r = 2.40) units, and the other two are group IV based corner-sharing ( r = 2.40) and edge-sharing ( r = 2.67) tetrahedral units. Remarkably, binary group V (P, As) sulfides exhibit IPs that are shifted to even a lower r than their selenide counterparts; a result that we trace to excess S n chains either partially (As–S) or completely (P–S) demixing from network backbone, in contrast to excess Se n chains forming part of the backbone in corresponding selenide glasses. In ternary chalcogenides of Ge with the group V elements (As, P), IPs of the sulfides are similar to their selenide counterparts, suggesting that presence of Ge serves to reign in the excess S n chain fragments back in the backbone as in their selenide counterparts.

The effect of NiO on the phase formation, thermo-physical properties and sealing behaviour of lithium zinc silicate glass–ceramics

We have prepared lithium zinc silicate (LZS) glasses of compositions (mol%) 17.83Li2O–17.73ZnO–(53.52 − x)SiO2–5.25Na2O–1.25P2O5–4.31B2O3–x-NiO, where 0.5 ≤ x ≤ 2.0, by the melt quench technique. The effect of NiO on the phase formation, thermo-physical properties and sealing behaviour of LZS glass–ceramics was studied using X-Ray diffraction (XRD), thermo-mechanical analysis (TMA) and microhardness (MH) measurements. It is found that NiO incorporation leads to a change in the role of ZnO from network modifier to intermediate oxide. The intermediate network forming Zn2+ ions would find it more difficult to diffuse and initiate the transformation of Li3Zn0.5SiO4 to Li2ZnSiO4. Thus Li3Zn0.5SiO4 is formed instead of Li2ZnSiO4 on addition of 2 mol% NiO. Scanning electron microscopy (SEM) and energy dispersive analysis of X-rays (EDAX) measurements at the glass–ceramic-to-metal interface reveal a change in the microstructure commensurate with the changing role of ZnO. Addition of NiO favoured interdiffusion of species at the interface leading to better sealing.

Proxy-Based Reference Picture Selection for Error Resilient Conversational Video in Mobile Networks

We propose a frame dependency management strategy for error robust transmission of conversational video in mobile networks. We consider an end-to-end video transmission scenario that involves both a wireless uplink as well as a wireless downlink plus some intermediate wireline network transmission. We also investigate the special cases of an end-to-end scenario where only a wireless uplink or a wireless downlink is present. We cope with packet loss on the downlink by retransmitting lost packets from the base station to the receiver for error recovery. Retransmissions are enabled by using fixed-distance reference picture selection during encoding with a prediction distance that corresponds to the round-trip time of the downlink combined with accelerated decoding. We deal with transmission errors on the uplink by sending acknowledgments and predicting the next frame to encode from those slices that have been correctly received by the base station. We show that these two separate approaches for uplink and downlink efficiently complement one another and the resulting end-to-end scheme is characterized by very low computational complexity. We compare our scheme to several state-of-the-art error resiliency approaches and report significant improvements.