Networked embedded software systems incorporate varying degrees of adaptation behavior to sustain their operations with acceptable quality of service (QoS), in the face of hostile external events. For instance, a highly agile cruise control system of a car may dynamically adjust its controller parameters to generate a higher-than-normal increase in torque when encountering a higher road elevation (relative to a normal controller). The paper formulates model-based assessment techniques to reason about how capable is a networked system in meeting its QoS specs in the presence of hostile external events. As a case study, we describe the QoS assessment of a multi-vehicle adaptive cruise control system.

Given cloud-based realization of a CDN service S (i.e., Content Distribution Network), QoS auditing captures the QoS violations that arise under various resource depletion and outage scenarios faced by S. Third-party control of the underlying cloud VM and storage nodes hosting the content (i.e., proxies of content server) raises the issue of reasoning about how well the CDN internal mechanisms are engineered to offer a required level of service to the application (i.e., low latency and overhead). We employ computational models of S to determine the optimal feasible proxy placements in the CDN topology and verify how close is the actual behavior of S to this 'gold standard'. Using declarative specifications, the QoS meta-data and CDN adaptation processes of S are externalized to enable a management module H reason about QoS violations. The paper elucidates the software and system engineering issues that arise in an external evaluation of the QoS provisioning of S by H.

Agent-level negotiations between an application and a cloud service provider (SP) need to consider two complexities: i) inaccuracy of computational models in capturing the behavior of a cloud-based system made up of diverse components and resources; and ii) disparity in the goals and priorities of an application and the SP in exercising a cloud-based system. The paper suggests the use of trial actions by an agent to learn about the system behavior and the opponent agent goals, as part of an effective negotiation strategy. The trial actions allow searching, alternations, and improvements on the system QoS space, towards a larger goal of optimal QoS control.

Given cloud-based realization of a distributed system S, QoS auditing enables risk analysis and accounting of SLA violations under various security threats and resource depletion faced by S. The problem of QoS failures and security infringements arises due to third-party control of the underlying cloud resources and components. Here, a major issue is to reason about how well the system internal mechanisms are engineered to offer a required level of service to the application. We employ computational models of S to determine the optimal feasible output trajectory and verify how close is the actual behavior of S to this trajectory. The less-than-100% trust between the various sub-systems of S necessitates our model-based analysis of the service behavior vis-a-vis the SLA negotiated with S. The paper describes the modeling techniques to analyze the dependability of such a cloud-based system.

We employ model-based software (MBS) techniques to assess the quality of adaptation in a network system S in the presence of uncontrollable external environment conditions. The lack of complete knowledge about the I/O behavior of S arises due to large dimensionality of the input parameter space and their interactions with the various components in S. The MBS techniques adapt the operations of S over multiple “observe-adapt” rounds and steer S towards a reference input specs. How close is the actual behavior to the reference input depicts the quality of adaptation by S. The paper provides a cyber-physical systems (CPS) based software structure to evaluate the non-functional attributes of the output behavior of S, and autonomically adjust the network and algorithm parameters therein for effective control of S. The paper corroborates our CPS-based approach with a case study of QoE-aware video rate adaptation over a bandwidth-limited network path.

Given the business mode of offering computing services to customers in a cloud setting, a major question arises: how good are the services of a cloud provider when compared to that of other providers. The paper attempts to answer the question by describing a methodology to measure the various cloud parameters (such as VM cycles and number of VM instances) at run-time and map them onto meaningful service-level attributes. The paper provides a concrete definition of the service attributes experienced by the client application: such as availability, agility, and elasticity, in terms of the underlying cloud infrastructure parameters (i.e., VM instances and network bandwidth). Since the IaaS parameters are hard to measure directly, we resort to a measurement methodology that maps the client-visible PaaS-layer service attributes onto the underlying IaaS parameters exported by the cloud provider. Our measurement methodology satisfies the requirements of cloud testing: stealthiness and non-intrusiveness, while minimizing the measurement overhead. The methodology is tested using a replicated image data store implementation on the PlanetLab.

Sensing errors may occur in the form of inaccurate and/or delayed reporting of system and/or external events. The sensing errors may skew the control actions performed by an adaptive network system S on its external environment - and hence contribute to control errors. Given that sensing and control errors do arise due to imprecise and incomplete knowledge about the computational models of S, a model-based treatment of sensing errors allows quantifying their impact on the control accuracy. Our paper describes a software cybernetics framework to deal with data sensing errors that arise during the operations of a complex network system.

We consider a wide-area video conferencing application where the video sources adapt their send rates according to the available bandwidth in the network paths. We advocate a QoE-aware cooperative rate control of the sources to relieve the congestion, instead of running multiple (independent) instances of a single-source adaptation algorithm in a QoE-oblivious manner and additively superposing their results. Our paper focuses on the architecture of such a QoE-aware video multicast system. Dove-tailed to the core functionality of rate adaptation is the session-layer configuration control mechanisms to deliver video to various end-user devices.

A system that is highly dependable under hostile conditions but whose dependability cannot be easily evaluated prior to the deployment of applications is less desirable than a system with lower but predictable dependability. This is because a decision-making on the deployment of high assurance systems is often based on a risk analysis of application failures. For systems implemented on a cloud, the problem of system certification assumes added importance because of third-party control of cloud resources and the attendant problems of faults, QoS degradations, and security violations. In this light, our paper focuses on: i) formulating metrics to quantify the dependability of cloud-based applications; and ii) identifying techniques to measure these metrics prior to deployment of applications.The paper treats system dependability as an application-level QoS for management purposes, and advocates a probabilistic evaluation of dependability. Our approach is corroborated by measurements on system-level prototypes and simulation analysis of system models in the face of hostile environment conditions. A case study of replicated data service anchored on cloud infrastructures is also described.

The paper deals with service abstractions that allow managing the dependability features of networked application systems.Component-level faults in the infrastructure are mapped on to degradations in the service quality space. Fault recovery mechanisms in a service-layer algorithm, if any, purport to maintain either the same level of service as in the no-fault case or a reduced level of service proportionate to the fault severity. Treating the occurrence of component faults as a normal case rather than as an exception, a virtualization mechanism provides a systematic mapping of faults on to canonical service quality spaces in a measurable way (e.g.,service availability and agility). Our paper describes the generic mechanisms and programming interfaces in the virtualization layer to manage the infrastructure in a flexible way that allows attaining an optimal service quality. Such service layer abstractions are useful for applications to in fuse adaptive fault-tolerance and quality assurance.

Given cloud-based realization of a distributed system S, QoS auditing enables risk analysis and accounting of SLA violations under various security threats and resource depletions faced by S. The problem of QoS failures and security infringements arises due to the third-party control of cloud resources and components that are used in realizing the application-oriented service exported by S. The less-than-100% trust between the various sub-systems of S is a major issue that necessitates a probabilistic analysis of the application behavior relative to the SLA negotiated with S. In this light, QoS auditing allows reasoning about how good the SLA is complied by S in the face of hostile environment conditions. The paper describes case studies of CDN and replicated web service realized on a cloud.

The paper describes a network service model where a service provider (SP) maintains multiple protocol modules to exercise the infrastructure resources under various environment conditions. Each protocol exhibits a certain degree of performance optimality and service resilience in distinct operating regions of the network infrastructure and the environment. During run-time, the SP selects one of the protocol modules that can meet the application-requested Quality of Service (QoS) obligation against the prevailing operating conditions. Under normal conditions when the external disturbances are benign (e.g., low packet loss in the network), an optimal usage of the network resources is important. Under adverse conditions however (e.g., heavy packet loss), a sustained access to the network service at some minimum acceptable level becomes more important than a resource-optimal service offering. Often, a resilient protocol incurs more resource usage to tackle the hostile environment conditions than a performance-conscious protocol tuned for the normal case operations. Accordingly, a protocol selection by the SP considers the trade-off between `service resilience' under extreme operating conditions and `resource optimality' under normal operations. Our model allows a dynamic switching from one protocol module to another at run-time based on the changing environment conditions.

The paper deals with a study of probabilistic methods to manage the dependability of a networked distributed system, in the presence of inaccuracies and partial knowledge of system models pertaining to failures. A distributed networked system (DIS) that collects raw data from sensors deployed in the field and delivers a composite data to an end-user is faced with two types of uncertainties: at 'information level' due to the multi-modal nature of raw data collected from the environment, and at 'control level' due to the incompleteness in knowledge about the application model. These have a compounded effect on the quality of fault-tolerance exhibited by a DIS. Based on service-layer abstractions, the paper identifies application-oriented metrics to quantify the quality of information flowing through a DIS. Even with imperfect information, the paper demonstrates how the high-level quality metrics and control algorithms enable achieving a reasonable degree of fault-tolerance in a probabilistic manner. A case study of replicated web services is also described.

We use the cyber-physical systems (CPS) framework to infuse intelligent adaptation behaviors in complex network systems to achieve dependability. The CPS framework is anchored on "intelligent physical worlds" (IPW) around which complex adaptation behaviors are built. An IPW is an embodiment of control software functions wrapped around the raw physical processes (e.g., servers, links, sensors, actuators), performing the core system activities while adapting its behavior to the changing environment conditions and user inputs. The IPW exhibits an intelligent behavior over a limited operating region of the system (in contrast with traditional models where the physical world is dumb). To perform over wide operating regions, the IPW interacts with an intelligent computational world (ICW) to patch itself with control parameters and procedures relevant in those changed conditions. The modular decomposition of an application into IPW and ICW lowers the design complexity of dependable network systems, and simplifies the system verification/testing.

Testing a distributed protocol P employed in a high assurance system involves the modeling of two distinct aspects: i) the intended goals G(P) to be satisfied by P, and ii) the external environment E(P) under which P operates. In complex distributed systems, a wide range of environment behaviors need to be captured by E(P) and/or diverse application requirements need to be stipulated in G(P). This model-based behavior generation is then used as a reference to compare with the actual behavior of a target system protocol. Our approach embodies two functional elements: first, an 'observer' at the service interface of P to the application that monitors the compliance to the conditions stipulated in G(P), and second, an 'injector' that subjects P to a variety of external disturbances as stipulated in E(P). The latter are either actual occurrences of system-level anomalies during an execution of P or computer-generated error conditions as aided by production rules that depict E(P). The modular decomposition that underlies our 3-tier approach (i.e., the observer and perturber modules interweaving with the target system module) enables a flexible testing of distributed protocols in various operating environments and under various application-level goals. The paper describes a language-based tool for generating events that drive the execution of P, and compare the observed behavior of P with a reference desired behavior.

The paper uses the cyber-physical systems (CPS) framework to model the intelligent adaptation behaviors in complex network systems. The CPS framework is anchored on intelligent physical worlds (IPW) around which complex adaptation behaviors are built. An IPW is an embodiment of control software functions wrapped around the raw physical processes (such as routers, links, hosts, and protocols), performing the core communication activities while adapting its behavior to the changing network conditions and user inputs. The IPW exhibits an intelligent behavior over a limited operating region of the network system, which is in contrast with the traditional models where the physical world is basically dumb. To perform over a wider range of network operating conditions, the IPW interacts with an intelligent computational world (ICW) to patch itself with suitable control parameters and rules & procedures relevant in those changed conditions. The modular decomposition of a network application into IPW and ICW lowers the design complexity of network systems, and simplifies the system verification/testing. The paper illuminates our CPS-based approach with a case study of adaptive video transport over a bandwidth-limited network.

In this paper, we provide a control-theoretic treatment of the resource allocations that adaptively occur in a QoS-aware network server system. Here, the target system being controlled is a logical service point that processes the transactions requested by clients using a resource infrastructure, with a goal of maximizing the revenues. Accurate management of resource allocations with a revenue-oriented goal is quite complex, due to the interactions among various transactions that dynamically share the resources in the system (such as server nodes, disks, content caches, and network bandwidth). So, we adopt an on-line monitor-and-control approach, aided by heuristics, that iteratively adjusts the resource allocation based on the observed transaction drop rate. We undertake a case study of end-to-end QoS-adaptive data transfer to illustrate the methodology. In terms of control theory, the bandwidth allocation and the packet loss rate constitute the system input and output respectively, with the heuristics-based bandwidth adjustment strategies incorporated in a controller along the feedback loop. The use of control theory allows offering predictable convergence properties of the QoS seen by applications, while maximizing the service provider revenues.

A key requirement for a distributed collaborative application is that all participants see the same copy of a shared window-object at different points in time (WYSIWIS). This is needed to maintain a cohesive view of the problem. We recognize that user actions are generated based on their understanding of the context of the problem, where context is defined in terms global view of the shared window object. However, it is noted that, actions generated from different sites in response to the same context may not be compatible with each other since execution of one action may change the context in such a way that other actions become irrelevant. This notion of compatibility of action is the crucial distinguishing factor between a distributed collaborative system and other distributed systems where independent actions are generally serializable. We propose a general framework that uses the concept of temporal epoch to implement a state-machine model of collaboration. We also explore the idea of `contextual integrity' as a pre-requisite for perceptual consistency among users.

The paper offers tree-structured connectivity between the peer entities over an infrastructure network as an abstraction to embody the application-oriented processing functions on peer-to-peer information flows. Tree reconfigurations are triggered when failures or security violations occur in the connectivity path between the peer nodes (e.g., increase in path delays, DOS attack on a node). Reconfigurations are also initiated when the application-level properties of information flows change (e.g., lower data accuracy for non-important contents). For Information Assurance settings, a tree structure elegantly supports the desired properties of separation and non-inference between the low-level network entities and the application entities. A connectivity management algorithm reconfigures a tree satisfying the desired topological properties (e.g., low geographic spread, low peer-to-peer distances). The algorithm provides redundancy in the connectivity paths along the temporal dimension (on top of a rich connectivity in the physical infrastructure - as in overlay networks). Application case studies of tree-structured connectivity with QoS and security goals are also described.

A real-time data collection system in sensor network settings requires dealing with failures in the network and in the external environment: such as data corruptions by malicious devices and message timeliness violations in the network. Functional replication is employed to deal with such failures, with voting among the replica devices to move a correct data to the end-user. The goal of our paper is to develop a voting system that dynamically adapts its internal mechanisms to deal with various types of failures. The paper presents the design issues, with considerations of protocol correctness and performance. A goal is to reduce the message overhead - and hence the power drain on wireless connected sensor devices.

In this paper, we study the impact of network-level message loss/delays on the performance experienced by the players in a real-time multi-player game. The study is conducted through the prism of consistency models of the game state (logically) shared among the players. We employ a weaker notion of consistency that encompasses: (i) the real-time persistence effects of player actions in the physical game world; and (ii) the state obsolescence caused by the loss/delay of player and game server messages in the network. Both (i) and (ii) impact the realism and cohesiveness of the game being played. The user-level performance measure, namely, the player-level lag, is studied for multi-player games under realistic settings of network communications and human participation. Towards this end, the paper formulated a formal model of multiplayer games and then studied the effects of network delays and human persistence on the player-level lags experienced. Sample games studied are the Soldier of Fortune and the RogueSpear. The methodology can be employed for studying the the effectiveness of human elements participating in the collaborative simulation of a physical world (such as war games).

In this paper, we provide an online monitor and control approach to manage bandwidth allocations on a shared network infrastructure in a QoS-aware application setting. Accurate management of bandwidth allocations with a goal of maximizing the revenues is quite complex, due to the interactions among various data flows that dynamically share the network bandwidth. We adopt a heuristics-aided control strategy that iteratively adjusts the bandwidth allocation based on the observed packet loss rate and delays. The paper first demonstrates the existence of a globally optimal operating point that minimizes the total bandwidth cost over a large number of flows while keeping the end-to-end packet delays within a prescribed limit. We then provide an algorithmic framework to search for this optimal point with an imprecise knowledge of the overall system model. Our optimization strategy for the QoS adaptation problem offers a unified management framework that considers both the bandwidth cost and the end-user QoS - and is hence useful for connectivity service providers.

The paper describes a policy-based model for cost-effective 'data connectivity' provisioning between session-level end-points. The connectivity provider (SP) may employ an architecture for end-to-end QoS control between data aggregation points. It involves: (i) maintaining multiple diffserv-type connections between end-points with parameterizable QoS differentiation between them, and (ii) admission control at end-points with intserv-type bandwidth management over connections, (ii) aggregates data flows with closely-similar QoS needs over a single end-to-end connection, (i) apportions the available infrastructure bandwidth between various end-to-end connections that carry (aggregated) data flows with distinct QoS levels. Flow aggregation over a connection allows reaping the statistical multiplexing gains in bandwidth, i.e., meets the SP's revenue incentives. Whereas, connection-level bandwidth allocation allows meeting the QoS needs of data flows, i.e., guarantees the end-user's utility. The management functions of SP monitor the changes and/or outages in network bandwidth in a dynamic setting, and maps them onto the connectivity costs incurred for Qos control. Our model allows installing policy functions at end-points that can make the connectivity provisioning cost-optimal.

The paper discusses the dynamics of bandwidth allocations that occur in an end-to-end QoS-aware data transport system over a LAN-MAN. Here, the controlled object is a logical connection that carries the application-level data flows over the service provider's data transport infrastructure. The goal is to allocate just-enough bandwidth, in order to maximize the service-provider revenues. Accurate control of bandwidth allocations with a revenue-oriented goal is quite complex, due to the interactions among various data flows that dynamically share the network capacity in the system. So, we employ an on-line monitor-and-control scheme, aided by heuristics, that iteratively adjusts the bandwidth allocation based on the observed packet loss rate. The bandwidth allocation is increased when the observed packet loss rate exceeds a user-specified threshold, with the increase being proportional to the loss rate (i.e., multiplicative increase). The bandwidth is decreased additively when the loss rate falls below a safe limit. We describe this MIAD-based bandwidth allocation scheme, and study its dynamics. When the LAN-MAN environments provide bandwidth capacities in a stable manner, our MIAD scheme can achieve an optimal bandwidth allocation (in the sense of user-level QoS and service-provider revenues).

The paper describes a protocol management architecture to enhance the performance of network service provisioning to clientele. In our model, a service provider (SP) maintains multiple protocol modules to exercise the infrastructure resources, with each protocol exhibiting a certain degree of cost optimality in distinct operating regions of the network infrastructure. During run-time, the SP selects one of the protocol modules that can meet the client-requested service obligation against the current resource-level operating conditions. The goal is to offer a sustained access to the service with a resource-optimal and QoS-compliant service offering. Accordingly, the protocol selection by SP needs to consider the tradeoff between `service sustainability' and `resource optimality' in various operating regions of the network. Often, a robust protocol is less efficient in infrastructure resource usage, and vice versa. Our model allows a `dynamic switching' from one protocol module to another at run-time based on the changing resource conditions. The paper describes `protocol switching' as an architectural foundation for building cost-effective network services. A sample network service: adaptive `content distribution network', is also described along with a simulation study

This paper describes the engineering of replica voting protocols for energy-efficiency in data delivery. The protocols employ 2-phase voting among replica processes to move a correct data from the external environment to the end-user(s) in a secure real-time application setting. The replicas may be wireless computation nodes deployed in a power-constrained environment. The cross-layer optimization techniques is employed to reduce the amount of network message exchanges and device processing cycles expended for data delivery to the user. This optimization can in turn reduce the battery energy consumption of wireless devices that participate in the voting protocol. Cross-layer design of a replica-based voting protocol is considered for deployment over IEEE- 802.11 networks. Furthermore, though the paper focused on replica voting as the application, the optimization techniques are useful in other application domains

End-to-end packet-pair probing has been used as one of the primary mechanisms to measure bottleneck capacity, cross-traffic intensity, and available bandwidth of end-to-end Internet paths. However, there has been lacking an understanding of what types of information about the path are sampled by packet-pairs and how they are encoded in the corresponding output dispersions. In this paper, we answer these questions by deriving the expression of individual output spacings (i.e., dispersions) in the context of multihop paths and general bursty cross-traffic arrivals. Based on these results, we examine stochastic properties of the sampled signals and derive the statistical mean of the output interpacket spacings as a recursive function of the input spacing. We show that the result is different from what has been obtained in prior work using fluid cross-traffic models and that this discrepancy has a significant adverse impact on the accuracy of packet-pair bandwidth measurement techniques

In a distributed embedded system, data items may be moved from one set of functional modules to another through a common buffer that is interposed between them over a network. Here, the computation modules disseminating data should agree on what data items are written into the remote buffer, to coordinate their actions on the external environment. Furthermore, a timely effect of these actions requires enforcing deadlines on the processing of data from the remote buffer. To meet these requirements, we provide a publish-subscribe style programming structure that enables a shared view of the processing on data in the buffer by various computation modules over prescribed time intervals. Our structure allows an explicit application-level control of the asynchrony and timing of information flow between various modules. Its use is illustrated with a case study of a target tracking application. In general, our programming structure can be useful in a variety of embedded software systems such as shared e-desks and collaborative Web services.

The paper deals with providing a specification methodology to enable flexible monitoring in large scale distributed information systems (DIS). The goal is to allow the monitoring of compliance to properties deemed as critical for functioning of a DIS. Since information attacks due to external intrusion and/or component failures often manifest as deviations from critical behaviors expected of a DIS, a violation of compliance to a specified critical property can be viewed as symptomatic of information attacks. Based on this principle, users can prescribe critical properties in the form of event predicates, which are Boolean conditions on the externally visible interface state distributed among computation nodes and can be detected by distributed algorithms. Our specification methodology manifests in two facets: i) designing a meta-language which allows the interface behavior of a system to be prescribed through possible occurrences of events in the external environment; and ii) incorporating the specification language into a programming interface that exposes the computation-level states symbolizing the occurrence of events. Any target application can then plugged-in to the generic monitor tool to test for its compliance to critical functionality requirements. The realization of monitor in distributed object-oriented programming languages (such as JAVA) reduces the software development costs of implementing a DIS: due to the ease of specification, development and maintenance of the DIS.

The multimedia information exchanges in distributed collaborative applications may be modeled, from communication system perspective, as the flow of media messages between user entities, to provide access to shared window objects on workstations. An example of shared window is a design document interactively reviewed by a team of engineers over a discussion session. A key requirement is that all users should see the same local copy of a shared window object at any given point in time (WYSIWIS), even though these users may reside in different workstations interconnected over a network. In this paper, we study the temporal ordering of real-time persistent messages as the basis for achieving WYSIWIS. The approach requires specifying the synchronization constraints on accessing shared window objects in the form of state-transition relations and timeliness needs when processing messages for object state updates. Non-conflicting actions can be processed in any sequence by user entities, thereby increasing overall system-level performance and user-level efficacy. The paper describes the salient features of our concurrency control model, and then identifies the basic protocol elements to realize the model. Applications are also described to demonstrate the viability of our approach.

Although packet-pair probing has been used as one of the primary mechanisms to measure bottleneck capacity, cross-traffic intensity, and available bandwidth of end-to-end Internet paths, there is still no conclusive answer as to what information about the path is contained in the output packet-pair dispersions and how it is encoded. In this paper, we address this issue by deriving closed-form expression of packet-pair dispersion in the context of a single-hop path and general bursty cross-traffic arrival. Under the assumptions of cross-traffic stationarity and ASTA sampling, we examine the statistical properties of the information encoded in inter-packet spacings and derive the asymptotic average of the output packet-pair dispersions as a closed-form function of the input dispersion. We show that this result is different from what was obtained in prior work using fluid cross-traffic models and that this discrepancy has a significant impact on the accuracy of packet-pair bandwidth estimation.

The paper employs majority voting among replica processes to move correct data from the environment to the end-user entity in a secure real-time application setting. The replicas may be wireless computation nodes deployed in a power-constrained environment, and the data may be quite large in size with nonnumeric and non-exact contents (e.g., imaging devices in a geographic terrain). The voting protocol is made energy-efficient by reducing the amount of data processing and network-level message exchanges required in delivering data to the user. Our optimization takes into account: (i) processing cycles expended in comparing data; (ii) amount of replica data movement required; (iii) number of control messages generated. We consider two types of voting protocol: a 'centralized' scheme where replica data are collected at a secure power-abundant site to carry out the data comparisons for voting; a 'decentralized' scheme where each replica compares its locally computed data with a candidate data to send its consent/dissent message to a central vote collating site. The paper develops a performance model that considers the (i)-(iii) tradeoffs to determine the energy consumption levels in the centralized and decentralized voting schemes. Our model allows the voting apparatus to select an energy-optimal scheme, given the timeliness parameters and the network and processing bandwidths.

In this paper, we take the sample-path approach in analyzing the asymptotic behavior of single-hop bandwidth estimation under bursty cross-traffic and show that these results are provably different from those observed under fluid models of prior work. This difference, which we call the probing bias, is one of the previously unknown factors that can cause measurement inaccuracies in available bandwidth estimation. We present an analytical formulation of "packet probing," based on which we derive several major properties of the probing bias. We then experimentally observe the probing bias and investigate its quantitative relationship to several deciding factors such as probing packet size, probing train length, and cross-traffic burstiness. Both our analytical and experimental results show that the probing bias vanishes as the packet-train length or packet size increases. The vanishing rate is decided by the burstiness of cross-traffic.

In a distributed collaborative application, a key requirement is that all users see the same copy of a shared window object at any given point in time (WYSIWIS). In this paper, we study 'user-assisted causal ordering' of messages as the basis for achieving WYSIWIS. The approach requires specifying the synchronization constraints on accessing shared window objects in the form of an order in which messages need to be processed and object state updated. The specifications are made available to the window subsystem based on the user-level knowledge about the actions on objects and the current (shared) object state. In contrast with the current approaches employing transaction models, our approach allows flexibility in the programming of collaboration-style applications, and offers increased levels of concurrency.

An infrastructure network often maintains nodes that participate in the distribution of video and/or content from a server to clients. To support these application sessions, the infrastructure network needs to maintain logical connectivity between the server and client entities, realized through transport-level path segments that interconnect one or more intermediate nodes. The connectivity can be viewed as an acyclic graph, i.e., a tree whose vertices include the nodes containing the client and server entities with edges set up over the intervening path segments. The paper deals with reconfiguration algorithms to handle the addition or removal of a vertex due to application-level actions like a client joining the distribution session or a server leaving. Dynamic network changes may also trigger the reconfiguration of a distribution path at the session level. With reconfiguration activities occurring asynchronously at the connectivity management level, coordination among these activities is required so that the reconfigured distribution path maintains the tree-structure (i.e., the connectivity graph remains acyclic). The paper provides an algorithmic framework that employs loosely synchronized time-stamp ordering to coordinate between multiple reconfiguration activities. Policy functions that prescribe constraints on the tree setups (e.g. minimum cost trees) can augment our algorithms to suit specific application domains. The usefulness of our framework in distribution path setups under resource constraints is also described. The framework is evaluated using discrete-event simulation.

The paper describes a management model for network service provisioning in which a service provider (SP) may maintain multiple protocol mechanisms to support its clientele. Client applications are presented with an abstract interface to these protocol mechanisms, through which service obligations may be prescribed. The management functions of SP track the changes and/or outages in network infrastructure resources in a highly dynamic setting, so that the service provisioning can be made resource-optimal. The management functions are based on a modular decomposition principle, which we refer to as 'virtual organization' of protocol modules. In this paradigm, a service may be viewed as provided by multiple protocol modules, with the SP selecting one of these modules and instantiating it with appropriate parameters to meet the service obligations. The model allows dynamic switching from one protocol module to another, based on a notion of cost associated with the infrastructure resource usage for a given level of service provisioning. The paper provides the functional mechanisms to realize the switching from one protocol to another for a client-transparent service provisioning. These mechanisms are based on our management-oriented studies of different types of information networks (such as content distribution networks and QoS-controlled data connections). Our model of service management offers a methodology for SPs to achieve cost-effectiveness for their service offerings.

In a distributed embedded system, one or more data items may be moved from one functional module to another through a common buffer that is interposed between them over a network. Here, the computation modules generating and/or disseminating data should agree on what data items are written into the remote buffer, to coordinate their actions on the external environment. Furthermore, a timely effect of these actions requires enforcing deadlines on the processing of data from the remote buffer. We encouch the functionalities of reaching agreement on the data processed through a shared buffer and meeting the time deadlines associated with this data in an abstract programming primitive: 'timed atomic write'. The primitive has a rich semantics in that it enables a consistent view of the processing on data in the buffer by various computation modules over prescribed time intervals. The primitive allows an explicit application-level control of the asynchrony and timing of information flow between various modules. The paper describes the structure and semantics of the 'atomic write' primitive, and illustrates its use in a 'sensor data fusion' application.

Current works on QOS-routing allow flow and QOS control to be exercised on multicast paths after an underlying routing protocol (e.g., DVMRP and CBT) sets up these paths. The 'QOS-control-after-routing' paradigm limits the ability of routers to provide cost-efficient routing (in a network-wide sense) because they can explore only those paths already set up even if more cost-efficient alternate paths exist. This motivates the need to integrate routing with flow and QOS control, wherein alternate paths can be considered to enable cost-efficient path setups. The paper proposes scalable IntServ-style mechanisms at the flow management and routing protocol levels.

From a graph-theoretical perspective, the problem of constructing multicast distribution paths, modeled as 'Steiner trees', is NP-complete. So, many heuristics-based algorithms are available to generate near-optimal trees. Typically, an algorithm first assigns the edge costs for all links, and then examines various candidate paths for interconnecting a given set of nodes. This strategy does not work well for the evolving multimedia application configurations (such as audio-video and image distributions) where it is often necessary to construct multiple distribution paths, viz., one per media data stream. This is because the overlapping of multiple tree segments over a common link forces the link cost to change, based on the delay and bandwidth characteristics of data streams flowing over these segments. Accordingly, algorithms that hitherto have assumed non-varying link costs during various phases of a run now need to take into account the variability of link costs as candidate trees with different levels of path overlapping are examined in a given run. In other words, as an algorithm runs by examining various paths, the link costs change. The paper embarks on a study of heuristics-based algorithms to tackle this 'modified Steiner tree' problem. The algorithms allow more cost-efficient routing of data than feasible otherwise with a classical treatment of the 'Steiner tree' problem.

Many multicast routing protocols consider only the unicast shortest paths between every pair of source and destination in a group (often based on the number of hops) to setup a multicast tree, such as DVMRP and CBT. Furthermore, they do not offer the ability to consider the resource savings possible in the presence of multiple data flows sharing common paths in a tree, such as CBT and PIM. These aspects hamper resource-efficient routing. In this light, our paper describes flow-based resource allocation control in multicast tree setups. Our proposed model has two aspects: (i) exploring alternate paths to connect a joining user to a tree, should the unicast shortest path not have sufficient bandwidth to sustain the new data flow; and (ii) downstream resource allocation to merge a new data flow with the on-going flows in a tree. (i). Requires 'multi-path connectivity' information at routers in the network, while (ii) requires flow aggregation information to be exchanged between on-tree routers. The model is evaluated by simulation studies. The proposed model can be useful in integrated service networks that contain both low and high bandwidth links, such as future Internet.

The authors examine the design and performance of LNTP, a connection-orientated transport-level protocol for local area networks (LANs) in the 4.2 BSD Unix operating system. Various measurements are taken, and LNTP's performance is compared to that of TCP/IP (ARPAnet's Transmission Control Protocol), a LHN protocol which is often used in LAN environments. The optimum values of various LNTP parameters are determined, some theoretically and others empirically. The theoretical results are compared to the experimentally observed values. It is concluded that LNTP does indeed outperform TCP/IP. However, due to the overhead of the non-LNTP specific protocol layers, this improvement is not as great as it might be. Nonetheless, LNTP proves itself to be a viable replacement for TCP/IP.>