Distributed Computing and Internet Technology

In critical infrastructures (CIs), different organizations must cooperate, while being mutually suspicious since they have different interests and can be in competition on some markets. Moreover, in most cases, there is no recognized authority that can impose global security rules to all participating organizations. In such a context, it is difficult to apply good security practices to the interconnected information systems that control the critical infrastructure. In this paper, we present the PolyOrBAC security framework, aimed at securing global infrastructures while preserving each participating organization’s autonomy. In this framework, each organization is able to protect its assets by defining its own security policy and enforcing it by its own security mechanisms, and the global infrastructure is protected by controlling and auditing all interactions between participating organizations. PolyOrBAC helps to satisfy the CII security requirements related to secure cooperation, autonomy and confidentiality, monitoring and audit, and scalability.

We enhance the notion of a computation of the classical theory of computing with the notion of interaction. In this way, we enhance a Turing machine as a model of computation to a Reactive Turing Machine that is an abstract model of a computer as it is used nowadays, always interacting with the user and the world.

In this paper we discuss our ongoing endeavour to apply notations and algorithms based on the π-calculus and its theories for the development of large-scale distributed systems. The execution of a large-scale distributed system consists of many structured conversations (or sessions) whose protocols can be clearly and accurately specified using a theory of types for the π-calculus, called session types. The proposed methodology promotes a formally founded, and highly structured, development framework for modelling and building distributed applications, from high-level models to design and implementation to static checking to runtime validation. At the centre of this methodology is a formal description language for representing protocols for interactions, called Scribble. We illustrate the usage and theoretical basis of this language through use cases from different application domains.

The technical capacities of service offers for e-government and e-participation have considerably progressed over the last years. Yet, the principles of good governance are still not well implemented, especially when it comes to policy development. Governments struggle to effectively apply innovative technologies in regards to providing open collaboration in policy formulation or to monitor and evaluate policy implementation. Through a recent initiative of the European Commission (EC), several research projects have been launched to address these challenges. This paper first investigates existing deficiencies in open government towards transparent policy development. Subsequently, an approach of a project funded by the EC is introduced to develop better ICT support for open collaboration in policy modeling. The approach combines existing e-participation tools, collaborative scenario generation and formal policy modeling to evaluate and explore policies via agent-based modeling (OCOPOMO - www.ocopomo.eu).

A linear process is a system of events and states related by an inner product, on which are defined the behaviorally motivated operations of tensor product or orthocurrence, sum or concurrence, sequence, and choice. Linear process algebra or LPA is the theory of this framework. LPA resembles Girard’s linear logic with the differences attributable to its focus on behavior instead of proof. As with MLL the multiplicative part can be construed via the Curry-Howard isomorphism as an enrichment of Boolean algebra. The additives cater for independent concurrency or parallel play. The traditional sequential operations of sequence and choice exploit process-specific state information catering for notions of transition and cancellation.

Reducing the time that a user has to occupy resources for completing cloud tasks can improve cloud efficiency and lower user cost. Such a time, called cloud time, consists of cloud deployment time and application running time. In this work we design jump-start cloud, under which an efficient cloud deployment scheme is proposed for minimizing cloud time. In particular, VM cloning based on disk image sharing has been implemented for fast VM and application deployment. For applications with heavy disk visits, the post-deployment quality of service (QoS) may suffer from image sharing and consequently, application running time will increase. To solve this problem, different image distribution schemes have been designed. We test jump-start cloud through a Hadoop based benchmark and MapReduce applications. Experiment studies show that our design saves application installation time and meanwhile, keeps application running time reasonably low, thus makes cloud time shorter.

As HPC (high performance computing) systems are extensively employed for heavy computational problems throughout heterogeneous environments, the scale and complexity of applications raises the issue of capacity planning. A cardinal aspect of efficiency is the job scheduler in any HPC systems. The job scheduling techniques can worsen or mitigate issues such as job starvation, increased queue time, and decreased system utilization. Since the impact of scheduling techniques is dependent on the workload of a supercomputer, this research proposes to analyze various scheduling disciplines on a given workload. By simulating HPC system, for any given workload, we can find the paradigm that yields the best performance, i.e. minimizing the wait time of jobs in the queue while maximizing resource utilization. Furthermore, given a fixed configuration of a HPC system, this research can be used to determine an appropriate workload that optimizes the system’s performance. The development and implementation of such complex simulation framework for HPC does not yet exist in HPC’s literature. The efficiency of the proposed simulation framework is illustrated through simulation results of performance measures such as average queuing time, average number of jobs in the queue, and system utilization. These results are verified by a developed mathematical model for job load characterization.

In object oriented databases, transactions may make simultaneous requests to do design time access and runtime access of resources. Concurrency control on the transactions can be implemented by using Multi granular lock models. Though several semantics based multi granular lock models have been proposed in the literature for object-oriented databases, they provide fine granularity of resources for runtime requests. However they have not fully utilized the semantics of object-oriented concepts to provide fine granularity of design time requests. In the proposed semantic based multi granular lock model, various dependencies of objects are exploited to exercise concurrency control. The dependencies among objects participating in the system can be inferred through their relationships such as inheritance, composition etc, with other objects participating in the domain. The proposed lock model uses these dependencies in defining lock modes for design time requests, which will provide fine granularity. It also utilizes these dependencies to provide better concurrency control for transactions modifying class relationships using Relationship Vectors (RV).

In the context of generating efficient, contention free schedules for inter-node communication through a switch fabric in cluster computing or data center type environments, all-to-all scheduling with equal sized data transfer requests has been studied in the literature [1,3,4]. In this paper, we propose a communication scheduling module (CSM) towards generating contention free communication schedules for many-to-many communication with arbitrary sized data. Towards this end, we propose three approximation algorithms - PST, LDT and SDT. From time to time, the CSM first generates a bipartite graph from the set of received requests, then determines which of these three algorithms gives the best approximation factor on this graph and finally executes that algorithm to generate a contention free schedule. Algorithm PST has a worst case run time of O( max (Δ|E|, |E|log(|E|))) and guarantees an approximation factor of 2H _2Δ− 1, where |E| is the number of edges in the bipartite graph, Δ is the maximum node degree of the bipartite graph and H _2Δ− 1 is the (2Δ− 1)-th harmonic number. LDT runs in O(|E|²) and has an approximation factor of 2(1 + τ), where τ is a constant defined as a guard band or pause time to eliminate the possibility of contention (in an apparently contention free schedule) caused by system jitter and synchronization inaccuracies between the nodes. SDT gives an approximation factor of 4log(w _max ) and has a worst case run time of O(Δ|E|log(w _max )), where w _max represents the longest communication time in a set of received requests.

Load balancing is very important and a common problem in distributed systems. The load balancing mechanism aims to fairly distribute load across the resources so as to optimize a given objective function. The objective can be system optimality which tries to minimize mean response time of all users or individual optimality which tries to minimize each user’s individual response time. The load balancing can be achieved either statically or dynamically. In this paper we review, competitive equilibrium (CE) approach for static load balancing and then propose recursive competitive equilibrium (RCE) approach for dynamic load balancing. A computer model is run to evaluate the performance of proposed RCE scheme with static scheme using Nash equilibrium (NE) approach and the static scheme using CE. The results show that static scheme using CE and dynamic scheme using RCE achieved both system optimality and individual optimality simultaneously, while NE scheme achieved only individual optimality. Moreover the performance of RCE scheme is higher than CE when communication overhead is less and almost same as CE scheme when the communication overhead is more.

We consider the problem of optimal routing in a multi-stage network of queues with constraints on queue lengths. We develop three algorithms for probabilistic routing for this problem using only the total end-to-end delays. These algorithms use the smoothed functional (SF) approach to optimize the routing probabilities. In our model all the queues are assumed to have constraints on the average queue length. We also propose a novel quasi-Newton based SF algorithm. Policies like Join Shortest Queue or Least Work Left work only for unconstrained routing. Besides assuming knowledge of the queue length at all the queues. If the only information available is the expected end-to-end delay as with our case such policies cannot be used. We also give simulation results showing the performance of the SF algorithms for this problem.

Given a set of sensors in the plane, the strong minimum energy topology ( SMET) problem is to assign transmit power to each sensor such that the resulting topology containing only bidirectional links is strongly connected. This problem is known to be NP-hard. As this problem is very much significant from application point of view, several heuristic algorithms have been proposed. In this paper, we propose a new incremental power greedy heuristic for SMET problem, called Kruskal-incremental power greedy heuristic. We compare Kruskal-incremental power greedy heuristic with Prim-incremental power greedy heuristic, one of the most popular heuristics available in the literature, through extensive simulation. The simulation results suggest that Kruskal-incremental power greedy heuristic outperforms on an average the Prim-incremental power greedy heuristic.

Wireless ad hoc network can be modeled as a unit disk graph (UDG) in which there is an edge between two nodes if and only if their Euclidean distance is at most one unit. The size of UDG is in O(n ²), where n is the number of network nodes. In the literature, the geometric spanners like Relative Neighborhood Graph (RNG), Gabriel Graph (GG), Delaunay Triangulation (Del), Planarized Localized Delaunay Triangulation (PLDel) and Yao Graph are proposed, which are sparse subgraphs of UDG. In this paper, we propose a hierarchy of geometric spanners called the k ^th order RNG (k-RNG), k ^th order GG (k-GG), k ^th order Del (k-Del), and k ^th order Yao (k-Yao) to reduce the spanning ratio and control topology, sparseness and connectivity. We have simulated these spanners and compared with the existing spanners. The simulation results show that the proposed spanners have better properties in terms of spanning ratio and connectivity by controlling topology and sparseness.

Localization of the nodes in a sensor network is a premier activity which influences the performance of the network. The data collected by a sensor node may become useless if the location of that node is not known. Sensor networks are mostly deployed in areas where manual positioning of the sensor nodes is not feasible, and the topology and size of the network also changes frequently. Therefore, the localization schemes developed for these kinds of network need to be self-configurable and adaptive to the changes. This paper presents a localization scheme for three dimensional wireless sensor networks that not only helps sensor nodes to self-localize, but, it is also able to verify the estimated location and re-estimate it, if needed. The sensor nodes estimate their positions based on the position information of the GPS enabled anchor nodes. Simulation results show that the proposed method gains heavily in terms of the accuracy of the estimated positions.

Currently, UDDI (Universal Description Discovery and Integration) is a standard for publishing and discovery of Web services. The UDDI search mechanism is based on functional properties and the consumer chooses a web service based on these properties. When functional properties are identical, additional distinguishing characteristics are needed to choose a service. Towards this, we propose to associate non functional properties, referred to as criteria, with a web service. A web service that has a list of criteria associated with it is termed as Decision Support Web Service (DSWS). In order to associate criteria, the UDDI is extended to X-UDDI which includes a new orange page containing the list of criteria and their description. We also add one more bag, criteria bag, which is used to define the attributes of the criteria. In order to publish and invoke a service on both, the functional properties and the criteria, we define new APIs.

In recent years more and more computer users are getting connected to the Internet. The explosive growth not only provides a wealth of opportunities for building online services, but also poses significant challenges. Handling an ever increasing number of users imposes a careful design of the system. Giving them real-time access to the data complicates even further the implementation of such a service. In this paper a real-time data storage architecture is presented that scales with ease to accommodate an increasing number of clients. The proposed architecture can handle not only predominant read operations, such as when using a traditional database with a caching layer, but also when most of the operations are write operations. By also prioritizing the access to data with respect to the user interactions with the system, the real-time performance of the system is enhanced.

Sharing of state resources like natural resources and developmental funds, needs to be inclusive as well as traceable. Further, for sustainability resource rights are to be enforced. For the purpose, a protocol for resource usages is proposed and shown that the protocol could be made adaptive to empower people by cooperation, collaboration and self help.

Tracking, extraction and recognition of text from video are important steps in building efficient indexing and retrieval systems for multimedia databases. We propose a top-down approach in which multiple cues are used for detecting text blocks in videos. In our proposed text tracking and extraction system, color and edge features are combined together to increase the precision of text detection. Our system detects both caption text as well as scene text of different font, size, color and intensity. Our objective is to detect both stationary text and moving text from complex color video. Extracted texts from videos are recognized using an OCR and stored in a database with relevant information. Such texts are used in future for retrieval of video clips based on any given keyword. This paper shows comparative result of RGB and HSV based approaches using different types of domain and shows the result of text tracking in different scenarios.

This paper demonstrates how to prevent or mitigate insider threats in relational databases. It shows how different order of accesses to the same data items may pose different levels of threat. Moreover, it states the conditions that are required to regard a data item as expired. In addition, it introduces the two different methods of executing insiders’ tasks, and how to prevent insider threat in those. The models presented in this paper organize accesses to data items in a particular sequence so that the availability of data items is maximized and the expected threat is minimized to the lowest level. Furthermore, it determines when to give an insider an incorrect but acceptable value of a risky data item in order to prevent a possible threat.

Privacy Preserving Publication has become one of the most prominent research topics in the recent years. Several techniques like k-anonymity, l-diversity and (α, k) anonymity were proposed to preserve privacy. Most of the published work focuses on anonymizing the microdata for preserving privacy and now the focus towards the verification of the anonymity levels of the microdata before publishing is the need of the day. Many publishers claim having anonymized the data. Verification of the claim on a large anonymized dataset is a herculean task. This paper focuses on providing simple approach for checking the anonymity levels for an anonymized dataset using frequent itemset generation. A GUI based tool named PRUDENT was developed to demonstrate the practicality of the solution. PRUDENT deals with numerical, categorical and multiple sensitive attributes. Results show that the algorithm is feasible and practical. A comparison with the existing methods is shown.

Image encryption schemes are important to ensure the security during transmission or storage. In this paper, chaos based image encryption scheme based on enhanced logistic map is proposed. The first stage consists of row and column rotation and XORing with first chaotic key. In the diffusion process, the pixel values are altered sequentially so that the changes made to a particular pixel depends on the accumulated effect of all the previous pixels. The operations include nonlinear diffusion using the second chaotic key and alternative zig-zag diffusion of adjacent pixels and XORing with the third chaotic key. The number of rounds in the steps are controlled by combination of pseudo random sequence and original image. The security and performance of the proposed image encryption technique have been analysed using statistical analysis, key space analysis, differential analysis and entropy analysis. The scheme possesses good performance in encryption speed and is suitable for real-time image encryption and transmission.

Insertion and deletion are considered to be the basic operations in Biology, more specifically in DNA processing and RNA editing. Based on these evolutionary transformations, a computing model has been formulated in formal language theory known as insertion-deletion systems. Since the biological macromolecules can be viewed as symbols, the gene sequences can be represented as strings. This suggests that the molecular representations can be theoretically analyzed if a biologically inspired computing model recognizes various bio-molecular structures like pseudoknot, hairpin, stem and loop, cloverleaf and dumbbell. In this paper, we introduce a simple grammar system that encompasses many bio-molecular structures including the above mentioned structures. This new grammar system is based on insertion-deletion and matrix grammar systems and is called Matrix insertion-deletion grammars. Finally, we discuss how the ambiguity levels defined for insertion-deletion grammar systems can be realized in bio-molecular structures, thus the ambiguity issues in gene sequences can be studied in terms of grammar systems.

In this paper we address sensor deployment problem to achieve different types of target coverage, viz; simple coverage, k-coverage and Q-coverage. Energy which is an important and scarce resource is not being optimally used if sensor nodes are randomly deployed in a region. This energy wastage can significantly be reduced if the deployment positions can be optimally computed. It is important to provide required coverage by keeping the required sensing range at minimum which will require less energy for sensing. We find out the optimal deployment positions in a 3-D terrain using Artificial Bee Colony (ABC) algorithm, which is based on swarm intelligence, and also compare the sensing range requirement for simple, k and Q-coverage problems. Experimental results reveal that for dense networks, the required sensing range does not increase in same proportion for increased value of k and increased value of average number of sensor nodes in Q for k-Coverage and Q-Coverage problems respectively. Sensitivity analysis is done to study the change in the required sensing range if the sensor nodes cannot be deployed exactly in the optimal positions. The analysis reveals that there is no significant change in the sensing range if the sensor nodes are deployed in near optimal positions.