Search

In Software-defined networks (SDNs), synchronization among network devices is an emergent phenomenon for enhancing the network scalability while preserving the centralized control. Owing to the fixed timeout values attributed to flow rules by the controller, flow table entries that are installed simultaneously are likely to timeout simultaneously and tend to trigger packet_in events at once during the upcoming sampling period. This phenomena in SDNs lead to serious performance and scalability issues when a large number of packet_in events are invoked simultaneously. In the proposed model, the orchestrating agent and the controller duo modulate an ensemble of networking devices from their heterogeneous frequency and incoherent nature of querying the controller into a conciliatory and systematic manner. The orchestrating agent modulates the control flows by altering timeouts, especially for long-lived TCP connections so that the devices do not overwhelm the controller with bursty requests during the future events. In addition, flow_mods to intermediate devices over a multi-hop path are scheduled based on the estimated flow setup latency to avoid redundant packet_ins. The proposed orchestrating agent is implemented in conjunction with the Floodlight controller. The experimental results show that apart from the influence of external delays majority of the flows were well-synchronized and the results show significant improvement in scalability and in the overall performance of the network. [ABSTRACT FROM AUTHOR]

Cloud interactive data-driven applications generate swarms of small TCP flows that compete for the small switch buffer space in data-center. Such applications require a small flow completion time (FCT) to be effective. Unfortunately, TCP is myopic with respect to the composite nature of application data. In addition it tends to artificially inflate the FCT of individual flows by several orders of magnitude, because of its Internet-centric design, that fixes the retransmission timeout (RTO) to be at least hundreds of milliseconds. To better understand this problem, in this paper, we use empirical measurements in a small data center testbed to study, at a microscopic level, the effects of various types of packet losses on TCP’s performance. In particular, we single out packet losses that impact the tail end of small flows, as well as bursty losses that span a significant fraction of small TCP congestion windows, and show a non-negligible effect of such losses on the FCT. Based on this, we propose the so-called, timely-retransmitted ACKs (or T-RACKs), a simple loss recovery mechanism that conceals the drawbacks of the long RTO even in the presence of heavy packet losses. Interestingly enough, T-RACKS achieves this transparently to TCP itself as it does not require any change to TCP in the tenant’s virtual machine (VM) or container. T-RACKs can be implemented as a software shim layer in the hypervisor between the VMs and the server’s NIC or in hardware as a networking function in a SmartNIC. Simulation and real testbed results show remarkable performance improvements. [ABSTRACT FROM AUTHOR]

An interactive application's overall pace of interaction is a combination of the user's pace and the system's pace, and if the system's pace is mismatched to the user's pace (e.g., timeouts or animations are too fast or slow for the user), usability and user experience can be impaired. Through a series of four studies, we investigated whether users prefer systems where the system's pace better matches their own pace. All of the studies used common drag-and-drop interactions with hierarchical folder widgets, in which a folder would expand when the cursor hovered over it for a timeout period. If the system pace in these interactions is too fast (i.e., the timeout is too short), then the user's performance and subjective experience is likely to be impaired because of unintended expansions; and if the system pace is too slow (i.e., the timeout is too long), then performance and experience could be impaired by unnecessary delay before folders expand. The first experiment was designed to validate the premise that fast-paced users prefer a fast system pace to a slow one (and the inverse for slow-paced users), and results confirmed this premise. The second study used the first experiment's data to look for measures of user pace that could enable automatic adaptation of system pace, and also examined whether participants adjusted their pace towards that of the system. The study found reliable measures of user pace and showed that participants do entrain to the system's pace. The third and fourth studies examined whether users would prefer a system that adapted its pace to the user over a system that used a static baseline pace. Results indicated that a majority of fast-paced users preferred the adaptive interface, but that slow-paced users generally preferred the static baseline interface. We discuss several design implications, including opportunities for systems to improve user experience for fast users by automatically adapting system pace to user pace. • Empirically shows that user preferences for system pace covary with user pace. • Analyses performance measures that can be used to automatically adapt system pace. • Shows that users converge their rate of interaction towards that of the system. • Shows that fast-paced users prefer an adaptive system pace to static pace. [ABSTRACT FROM AUTHOR]

We define a process calculus to describe multi-agent systems with timeouts for communication and mobility able to handle knowledge. The knowledge of an agent is represented as sets of trees whose nodes carry information; it is used to decide the interactions with other agents. The evolution of the system with exchanges of knowledge between agents is presented by the operational semantics, capturing the concurrent executions by a multiset of actions in a labelled transition system. Several results concerning the relationship between the agents and their knowledge are presented. We introduce and study some specific behavioural equivalences in multi-agent systems, including a knowledge equivalence able to distinguish two systems based on the interaction of the agents with their local knowledge.

The purpose of this study was to analyse the variables (lost rallies and score difference) that determine the timeout effect (positive or no effect) in volleyball, in balanced and unbalanced sets. 232 timeouts, requested by the coaches of 66 male and female teams participating in the Spanish Championship in the Under-14 and Under-16 categories, were analysed. The variables considered in this study were timeout effects, lost rallies and score differences. To analyse the timeout effect, a binary logistic regression model was applied. The results of this model show that, in balanced sets, the variables that predict the timeout effect are the number of rallies (≤2 lost rallies) and the score difference (2–3 points), whilst in unbalanced sets, and the variable that predicts the timeout effect is the number of lost rallies (3 lost rallies). These results show the importance of bearing these variables in mind when timeouts are managed and requested by coaches, in order to optimise the team’s performance.

The purpose of this study was to analyse the variables (lost rallies and score difference) that determine the timeout effect (positive or no effect) in volleyball, in balanced and unbalanced sets. 232 timeouts, requested by the coaches of 66 male and female teams participating in the Spanish Championship in the Under-14 and Under-16 categories, were analysed. The variables considered in this study were timeout effects, lost rallies and score differences. To analyse the timeout effect, a binary logistic regression model was applied. The results of this model show that, in balanced sets, the variables that predict the timeout effect are the number of rallies (≤2 lost rallies) and the score difference (2–3 points), whilst in unbalanced sets, and the variable that predicts the timeout effect is the number of lost rallies (3 lost rallies). These results show the importance of bearing these variables in mind when timeouts are managed and requested by coaches, in order to optimise the team's performance. [ABSTRACT FROM AUTHOR]

In the management of team sport competition, timeouts are one of the main tools of the coach to stop the game. It could be used on changing trends or the rhythm of the game, and also on providing relevant information to the players. So, it may have an important effect in sporting success. The main objective of the research was to analyse the use of timeouts (TO) in U-19 male volleyball teams, as well as the effect of the game on then, depending on the match score (winning-balanced, or losing). 171 timeouts, requested by coaches of the 21 teams participating in the U-19 male Spanish Championship 2012, were analysed. The variables considered in this study were: match score (TO requested when winning-balanced score or when losing); score difference (0-1 points, 2-3 points, 4 or more points); lost rallies (0-1 rallies, 2 rallies, ≥ 3 rallies); game period (initial moment of the set, final moment of the set); timeout effect (positive timeout effect and time out with no effect). The results showed that variables significantly associated with the TO request with different score (winning-balanced, losing) were: score difference, game period and timeout effect. Specifically, contributed positively to the association: TO request with winning-balanced score with the score difference of 0-1 points, with the final moments of the set and with the positive timeout effect; TO request with a losing score with the score difference of 4 or more points, in the initial moments of the set and with no effect of the timeout. This results could be taken into account in the process of team management in competition, in volleyball formative stages. [ABSTRACT FROM AUTHOR]

For addressing problems such as long delays, latency fluctuations, and frequent timeouts in conventional Transmission Control Protocol (TCP) in a data center environment, Data Center TCP (DCTCP) has been proposed as a TCP replacement to satisfy the requirements of data center networks. It is gaining more popularity in academic as well as industry areas due to its performance in terms of high throughput and low latency, and is widely deployed in data centers. However, according to the recent research about the performance of DCTCP, authors have found that most times the sender’s congestion window reduces to one segment, which results in timeouts. In addition, the authors observed that the nonlinear marking mechanism of DCTCP causes severe queue oscillation, which results in low throughput. To address the above issues of DCTCP, we propose multiple congestion points using double threshold and congestion reaction using window adjustment (DT-CWA) mechanisms for improving the performance of DCTCP by reducing the number of timeouts. The results of a series of simulations in a typical data center network topology using Qualnet network simulator, the most widely used network simulator, demonstrate that the proposed window-based solution can significantly reduce the timeouts and noticeably improves the throughput compared to DCTCP under various network conditions.

The problem of TCP incast in data centers attracts a lot of attention in our research community. TCP incast is a catastrophic throughput collapse that occurs when multiple senders transmitting TCP data simultaneously to a single aggregator. Based on several experiments, researchers found that TCP timeouts are the primary cause of incast problem. Particularly, timeouts due to insufficient duplicate acknowledgments is unavoidable when at least one of the last three segments is lost from the tail of a window. As a result, this type of timeouts should be avoided to improve the goodput of TCP in data center networks. A few attempts have been made to reduce timeouts, but still the problem is not completely solved especially in the case of timeouts due to insufficient duplicate acknowledgments. In this paper, we present an efficient TCP fast retransmission approach, called TCP-EFR, which is capable to reduce TCP timeouts due to lack of duplicate acknowledgments which is caused by the loss of packets from the tail of a window in data center networks. TCP-EFR makes changes in the fast retransmission and recovery algorithm of TCP by using the congestion signal mechanism of DCTCP based on instantaneous queue length. In addition, TCP-EFR controls the sending rate for avoiding the overflow of switch buffer in order to reduce the loss of packets. The results of a series of simulations in single as well as multiple bottleneck topologies using qualnet 4.5 demonstrates that TCP-EFR can significantly reduce the timeouts due to inadequate duplicate acknowledgments and noticeably improves the performance compared to DCTCP, ICTCP and TCP in terms of goodput, accuracy and stability under various network conditions. [ABSTRACT FROM AUTHOR]

Abstract: The bandwidth is one of the most precious resources in Mobile Ad Hoc Networks (MANETs); their network protocols should therefore be able to achieve an increase in efficiency of bandwidth utilisation. TCP, which implements flow control to regulate the network traffic, plays a significant role in shaping the workload of the network. However, spurious TCP retransmissions can increase the link utilisation, consequently limiting the amount of spare bandwidth available as extensively reported in the literature. This problem is exacerbated by the presence of high priority traffic and traffic differentiation introduced in IEEE 802.11e. In this paper, we focus on the optimisation of TCP for resource efficiency, by limiting the number of contention-induced retransmissions. Using a cross-layer technique, TCP can be notified of the medium contention state and adjusts its retransmission timer accordingly. Our simulation results demonstrate the effectiveness of our proposal in various scenarios. A 15% gain in resource utilisation is obtained, fewer packets are retransmitted thus reducing the overall power consumption and TCP goodput is improved. [Copyright &y& Elsevier]

The aim of the present study was to assess the effects of game timeouts on basketball teams' offensive and defensive performances according to momentary differences in score and game period. The sample consisted of 144 timeouts registered during 18 basketball games randomly selected from the 2007 European Basketball Championship (Spain). For each timeout, five ball possessions were registered before (n=493) and after the timeout (n=475). The offensive and defensive efficiencies were registered across the first 35 min and last 5 min of games. A k-means cluster analysis classified the timeouts according to momentary score status as follows: losing (-10 to -3 points), balanced (-2 to 3 points), and winning (4 to 10 points). Repeated-measures analysis of variance identified statistically significant main effects between pre and post timeout offensive and defensive values. Chi-square analysis of game period identified a higher percentage of timeouts called during the last 5 min of a game compared with the first 35 min (64.9±9.1% vs. 35.1±10.3%; χ2=5.4, P<0.05). Results showed higher post timeout offensive and defensive performances. No other effect or interaction was found for defensive performances. Offensive performances were better in the last 5 min of games, with the least differences when in balanced situations and greater differences when in winning situations. Results also showed one interaction between timeouts and momentary differences in score, with increased values when in losing and balanced situations but decreased values when in winning situations. Overall, the results suggest that coaches should examine offensive and defensive performances according to game period and differences in score when considering whether to call a timeout. [ABSTRACT FROM AUTHOR]

A link model-driven approach toward transmission control protocol (TCP) performance over a wireless link is presented. TCP packet loss behavior is derived from an underlying two-state continuous time Markov model. The approach presented here is (to our knowledge) the first that simultaneously considers (1) variability of the round-trip delay due to buffer queueing; (2) independent and nonindependent (bursty) link errors; (3) TCP packet loss due to both buffer overflow and channel errors; and (4) the two modes of TCP packet loss detection (duplicate acknowledgments and timeouts). The analytical results are validated against simulations using the ns-2 simulator for a wide range of parameters; slow and fast fading links; small and large link bandwidth-delay products. For channels with memory, an empirical rule is presented for categorizing the impact of channel dynamics (fading rate) on TCP performance. [ABSTRACT FROM PUBLISHER]

We define a process calculus to describe multi-agent systems with timeouts for communication and mobility able to handle knowledge. The knowledge of an agent is represented as sets of trees whose nodes carry information; it is used to decide the interactions with other agents. The evolution of the system with exchanges of knowledge between agents is presented by the operational semantics, capturing the concurrent executions by a multiset of actions in a labelled transition system. Several results concerning the relationship between the agents and their knowledge are presented. We introduce and study some specific behavioural equivalences in multi-agent systems, including a knowledge equivalence able to distinguish two systems based on the interaction of the agents with their local knowledge. [ABSTRACT FROM AUTHOR]

For addressing problems such as long delays, latency fluctuations, and frequent timeouts in conventional Transmission Control Protocol (TCP) in a data center environment, Data Center TCP (DCTCP) has been proposed as a TCP replacement to satisfy the requirements of data center networks. It is gaining more popularity in academic as well as industry areas due to its performance in terms of high throughput and low latency, and is widely deployed in data centers. However, according to the recent research about the performance of DCTCP, authors have found that most times the sender’s congestion window reduces to one segment, which results in timeouts. In addition, the authors observed that the nonlinear marking mechanism of DCTCP causes severe queue oscillation, which results in low throughput. To address the above issues of DCTCP, we propose multiple congestion points using double threshold and congestion reaction using window adjustment (DT-CWA) mechanisms for improving the performance of DCTCP by reducing the number of timeouts. The results of a series of simulations in a typical data center network topology using Qualnet network simulator, the most widely used network simulator, demonstrate that the proposed window-based solution can significantly reduce the timeouts and noticeably improves the throughput compared to DCTCP under various network conditions. [ABSTRACT FROM AUTHOR]