Real time resource usage prediction is an important part of resource provisioning in a cloud data centre. As cloud workloads vary dynamically, effective resource provisioning requires prediction of future resource usage trends. The problem is highly complicated because of highly time varying nature of cloud resource workloads. Training the future resource usage prediction models once, using a fixed set of observations is not sufficient to capture the variability in cloud workloads. In this work, we propose to use gradient descent (GD) and Levenberg-Marquardt (LM) adaptation algorithms for dynamic adaptation of resource utilization prediction models. We also propose a novel sparse framework for fast online adaptation of resource usage prediction models. We propose to analyze different algorithms such as ℓ1regularization, ℓ2regularization, optimal brain damage (OBD), optimal brain surgeon (OBS) for introducing sparsity. The proposed sparse framework for online adaptation of multivariate resource usage prediction models is validated for CPU usage prediction in the Google cluster trace and PlanetLab workload trace. A comparative analysis of different sparse frameworks shows that OBD-based LM adaptation algorithm performs better than other frameworks for online multivariate resource usage prediction in a cloud.

Abstract Current access control models have limitations in scenarios where digital objects are decomposed into smaller objects before sharing. We propose an access control model (digital object‐based access model) for such scenarios involving object decomposition. Our model applies to the generic class of objects composing of documents, videos, and audio. We use temporal logic of actions to formally specify access policies in our model and verify different properties. Using mathematical modelling and simulation, we compare the usability and performance of our model in comparison with the widely used traditional access control model and role‐based access control model. With business process outsourcing as an example, we demonstrate that our model also yields a considerable reduction in the number of access permissions to be managed by an administrator as compared with other two models. Simulations show an improvement in the time required for access authorization in our model as compared with these two models. These improvements exhibited by our model become more prominent with the increase in the number of users and objects. Copyright © 2015 John Wiley & Sons, Ltd.

Synchronous Digital Hierarchy (SDH) and Synchronous Optical Network (SONET) are time-division multiplexing technologies widely used in transport networks to provide bandwidth services. Dynamic service provisioning requires the use of online algorithms that automatically compute the path to be taken to satisfy the given service request. SDH/SONET multiplexing structure imposes restrictions on the allocation of bandwidth. Bandwidth services mostly require protection so that they are resilient to failures in the network. Two types of protection techniques—namely, dedicated protection and shared protection—are commonly used to protect bandwidth. Also, SDH/SONET provides standard protection mechanisms like linear MSP/APS, MS-SPRing/BLSR and SNCP/UPSR. In shared protection, complete path information for all the existing connections has to be maintained to maximize sharing for future service requests. This information will be typically very large. To overcome this problem, a minimum information scenario is proposed as a tradeoff between maintaining less information and maximizing sharing. In this work, heuristic algorithms are proposed for dedicated and shared protection that take into account the above-mentioned factors. These algorithms use the path computation algorithms for SDH networks proposed in a previous work of the authors, which compute paths for service requests without considering protection. The performance is evaluated for different values for the relative weight of sharing for the complete and minimum information scenarios. An extension to these algorithms for satisfying bandwidth requests that use various inverse multiplexing techniques is also proposed.

Most of the search engine optimization techniques attempt to predict users interest by learning from the past information collected from different sources. But, a user's current interest often depends on many factors which are not captured in the past information. In this paper, we attempt to identify user's current interest in real time from the information provided by the user in the current query session. By identifying user's interest in real time, the engine could adapt differently to different users in real time. Experimental verification indicates that our approach is encouraging for short queries

Synchronous digital hierarchy (SDH) is a time division multiplexing technology widely used in transport networks to provide bandwidth services. Dynamic service provisioning refers to the arrival of service requests one-by-one randomly with no prior information on future requests. This requires the use of on-line algorithms which automatically compute the path to be taken to satisfy the given service request. This problem involves a tradeoff between minimizing the number of requests that are rejected and minimizing the total bandwidth that is utilized. Many earlier works have addressed path computation algorithms, but they treat each link as having some integer units of bandwidth. They do not take into account the multiplexing structure defined by SDH which imposes restrictions on the allocation of bandwidth and the fact that higher order trails (logical connections) have to be established to support any bandwidth requirement. In this work, these factors are considered in the path computation algorithms. The network is treated as a graph containing physical links and logical trails and weights are assigned to them before computing a path with the least cost. Weights are assigned such that the trails are given higher preference to physical links so that existing trails are used wherever possible. This avoids unnecessary creation of new trails. The performance is evaluated for different values of weights. An improvement in the form of dynamically adjusting the weights of links and trails is done and its performance is shown to be better than having constant weights.

Several efficient reliable broadcast protocols have been developed and implemented on broadcast LANs. To enhance the amount of bandwidth available per node on a shared media LAN, such as 10 Mb/s Ethernet, there is a rapidly growing usage of LAN switches to interconnect several independent LAN segments to form an Extended LAN (E-LAN). The existing reliable broadcast protocols may be inefficient in E-LANs because the LAN switch in an E-LAN transmits both broadcast or multicast traffic on all the segments. However, in many group applications group members may be present only on few segments. In this paper, we propose some reliable broadcast protocols (REP) that are designed to work efficiently in E-LANs. This is achieved by added functionality to a programmable switch. We develop simple deterministic models for performance of RBPs. We show that our switched RBPs are more efficient than the existing RBPs in an E-LAN.

Using simulation, a network-independent framework compares the performance of contention-based Ethernet and two contention-free round-robin schemes, namely Expressnet and the IEEE 802.4 token bus. Two priority mechanisms for voice/data traffic on round-robin networks are studied: the alternating-rounds mechanism of the Expressnet, and the token rotation timer mechanism of the token bus, which restricts access rights based on the time taken for a token to make one round. It is shown that the deterministic schemes almost always perform better than the contention-based scheme. Design issues such as the choice of minimum voice packet length, priority parameters, and voice encoding rate are investigated. An important aspect that is noted is the accurate characterization of performance over a wide region of the design space of voice/data networks.>

A simulation study is presented in several distributions of stations on a linear-bus Ethernet. Aggregate performance is shown to depend on the distribution of stations. Individual station performance varies with the location of the station. Unbalanced distributions can lead to large performance differences between individual stations, with isolated stations achieving relatively poor performance compared to the average. The effects of access protocol parameters such as the number of buffers per station and the retransmission algorithm are examined. A modification of the standard retransmission algorithm is presented that enables a higher throughput to be achieved at high load. The results are compared to the predictions of theoretical models, and the applicability of the models to finite-population Ethernets is examined.>

In the past few years, local area networks have come into widespread use for the interconnection of computers. Together with the trend towards digital transmission in voice telephony, this has spurred interest in integrated voice/data networks. The Expressnet, an implicit-token round-robin scheme using unidirectional busses, achieves high performance even at bandwidths of 100 Mb/s. Other features that make the protocol attractive for voice/data traffic are bounded delays and priorities. The latter is achieved by devoting alternate rounds to one or the other of the two traffic types. By the use of accurate simulation, the performance of the Expressnet with voice/data traffic is characterized. It is shown that the Expressnet satisfies the real-time constraints of voice traffic adequately even at bandwidths of 100 Mb/s. Data traffic is able to effectively utilize bandwidth unused by voice traffic. The trade-offs in the alternating round priority mechanism are quantified. Loss of voice samples under overload is shown to occur regularly in small, frequent clips, subjectively preferable to irregular clips. In a comparison of the Expressnet, the contention-based Ethernet and the round-robin Token Bus protocols, the two round-robin protocols are found to perform better than the Ethernet under heavy load owing to the more deterministic mode of operation. The comparison of the two round-robin protocols highlights the importance of minimizing scheduling overhead at high bandwidths.

Local computer networks are increasing in popularity for the interconnection of computers for a variety of applications. One such network that has been implemented on a large scale is the Ethernet. This paper describes an experimental performance evaluation of a 3 and a 10 Mb/s Ethernet. The effects of varying packet length and transmission speed on throughput, mean delay and delay distribution are quantified. The protocols are seen to be fair and stable. These measurements span the range from the region of high performance of the CSMA/CD protocol to the upper limits of its utility where performance is degraded. The measurements are compared to the predictions of existing analytical models. The correlation is found to range from good to poor, with more sophisticated models yielding better results than a simple one.

The problem of modelling and analysing the performance of software structures for distributed computer systems is addressed. A modelling technique is proposed which has many striking analogies with current techniques for evaluating hardware systems and yet focuses attention on the system software. The analysis of such models will draw heavily on the substantial work already done in the analysis of hardware models.