Dynamic Distributed Environment Research Articles

Resilient multi-agent RL: introducing DQ-RTS for distributed environments with data loss

This paper proposes DQ-RTS, a novel decentralized Multi-Agent Reinforcement Learning algorithm designed to address challenges posed by non-ideal communication and a varying number of agents in distributed environments. DQ-RTS incorporates an optimized communication protocol to mitigate data loss between agents. A comparative analysis between DQ-RTS and its decentralized counterpart Q-RTS, or Q-learning for Real-Time Swarms, demonstrates the superior convergence speed of DQ-RTS, achieving a remarkable speed-up factor ranging from 1.6 to 2.7 in scenarios with non-ideal communication. Moreover, DQ-RTS exhibits robustness by maintaining performance even when the agent population fluctuates, making it well-suited for applications requiring adaptable agent numbers over time. Additionally, extensive experiments conducted on various benchmark tasks validate the scalability and effectiveness of DQ-RTS, further establishing its potential as a practical solution for resilient Multi-Agent Reinforcement Learning in dynamic distributed environments.

Scalable Access Control Scheme of Internet of Things Based on Blockchain

With the development of Internet of things (IoT) technology, a large-scale, heterogeneous and dynamic distributed IoT environment has gradually formed between different IoTs. In order to solve the scalability problem of restricted device access management in the Internet of things, a distributed access control system model of the Internet of things based on blockchain technology is proposed. The system model adopts a single smart contract, which simplifies the whole process in the blockchain network and reduces the communication overhead between nodes. According to the simulation results and evaluation, it is proved that the solution has good scalability.

PAutoBotCatcher: A blockchain-based privacy-preserving botnet detector for Internet of Things

Botnets have become a major threat in the Internet of Things (IoT) landscape, due to the damages that these sets of compromised IoT devices may cause. To increase their attacks’ success, modern botnets are designed in a distributed manner, following a P2P structure. Recently, several botnet detection solutions have been proposed. Among them, community behavior analysis solutions seem to be promising because of their high detection accuracy. However, such solutions are not optimized for real life scenarios since they only run in a static mode, that is, reading all network traffic at once. As such, they do not support real-time data analysis. In order to handle such issue, these solutions should run in a dynamic distributed environment where different actors participate in the detection process. However, this collaborative environment brings up the issue of trust among the actors.To address this issue, in this paper, we present PAutoBotCatcher, a dynamic botnet detection framework based on community behavior analysis among peers managed by different actors. PAutoBotCatcher leverages on blockchain to ensure immutability and transparency among all actors. To optimize continuous detection while keeping good accuracy, we design a set of optimization techniques, such as caching detection’s output and pre-processing the shared network traffic. In addition, we leverage on different privacy-preserving techniques to protect devices from re-identification during the botnet detection process. We have extensively tested our solution to show its effectiveness and to demonstrate that blockchain is a good solution for dynamic botnet detection.

A Novel Gmtds Algorithm for Dynamic Management of Transaction under Different Workload Condition

In today’s scenario large enterprise world spread across different locations, continents or having a diverse presence over the globe, where data is enormous and handling such large data over distributed computing becomes critical in real-time database system. The transaction management system for the distributed environment must ensure that the sequence of updates in the stable warehouses in different locations is confirmed or cancelled safely as a single complete unit of work. Working with Real-Time Database System (RTDBS) and that to on a distributed computing system is a tough task. When we work with distributed environment over larger database, we need to take care of the transaction time period as well as number of transactions that are actually executed (committed) and number of transactions fail. The application on dynamic RTDBS becomes more complex when certain deadlines need to be completed. In this paper, we had carried out the test of CRUD (Create, Read, Update, and Delete) operation on transactional Real-time databases in real time dynamic distributed environment by using the existing EDF, GEDF algorithm and we had compared these algorithms with our proposed GMTDS algorithm in standalone and distributed environment with a dynamic self adaptive approach for management of transactions.

Hybrid evolutionary workflow scheduling algorithm for dynamic heterogeneous distributed computational environment

The optimal workflow scheduling is one of the most important issues in heterogeneous distributed computational environments. Existing heuristic and evolutionary scheduling algorithms have their advantages and disadvantages. In this work we propose a hybrid algorithm based on heuristic methods and genetic algorithm (GA) that combines best characteristics of both approaches. We propose heuristic algorithm called Linewise Earliest Finish Time (LEFT) as an alternative for HEFT in initial population generation for GA. We also experimentally show efficiency of described hybrid schemas GAHEFT, GALEFT, GACH for traditional workflow scheduling as well as for variable workload in dynamically changing heterogeneous computational environment.

A bio-inspired hybrid algorithm for effective load balancing in cloud computing

Effective scheduling and load balancing are the key challenges of cloud computing technology. A prominent task scheduler must be adaptable to the dynamic distributed environment and to the job scheduling policy based upon the workload. In this research, a novel hybrid algorithm is developed for balancing the load among cloud nodes by hybridising the existing ant colony optimisation (ACO) algorithm and priority-based artificial bee colony (ABC) algorithm where ACO is used for balancing the workload and ABC is used to optimise the resource scheduling from the prospect of cloud computing systems. The proposed technique is tested to show that proposed approach minimise the average response time, average data centre processing time and total processing cost to serve the user requests.

A bio-inspired hybrid algorithm for effective load balancing in cloud computing

Effective scheduling and load balancing are the key challenges of cloud computing technology. A prominent task scheduler must be adaptable to the dynamic distributed environment and to the job scheduling policy based upon the workload. In this research, a novel hybrid algorithm is developed for balancing the load among cloud nodes by hybridising the existing ant colony optimisation (ACO) algorithm and priority-based artificial bee colony (ABC) algorithm where ACO is used for balancing the workload and ABC is used to optimise the resource scheduling from the prospect of cloud computing systems. The proposed technique is tested to show that proposed approach minimise the average response time, average data centre processing time and total processing cost to serve the user requests.

Data mining based multi-level aggregate service planning for cloud manufacturing

Cloud manufacturing (CMfg) promotes a dynamic distributed manufacturing environment by connecting the service providers and manages them in a centralized way. Due to the distinct production capabilities, the service providers tend to be delegated services of different granularities. Meanwhile, users of different types may be after services of different granularities. A traditional aggregate production planning method is often incapable of dealing with type of problems. For this reason, a multi-level aggregate service planning (MASP) methodology is proposed. The MASP service hierarchy is presented, which integrates the services of different granularities into a layered structure. Based on this structure, one of data mining technologies named time series is introduced to provide dynamic forecast for each layer. In this way, MASP can not only deal with the services of multi-granularity, but also meet the requirements of all related service providers irrespective of their manufacturing capabilities. A case study has been carried out, showing how MASP can be applied in a CMfg environment. The results of the prediction are considered reliable as the order of magnitude of the production for each service layer is much greater than that of the corresponding mean forecast error.

Adaptive Spectrum Management of Cognitive Radio in Intelligent Transportation System

With the rapid development of urban rail transit, the demand of the public in rail transportationto take real-time, reliable and efficient wireless access services, has become the focus of mobilebroadband communications. Wireless cognitive radio (CR) over urban rail transit is a newly emergingparadigm that attempts to opportunistically transmit in licensed frequencies, without affecting the preassignedusers of these bands. To enable this functionality, such a radio must predict its operationalparameters, such as transmit power and spectrum. These tasks, collectively called spectrum management,is difficult to achieve in a dynamic distributed environment, in which CR users may only takelocal decisions, and react to the environmental changes. In this paper, we propose a reinforcementlearning based approach for spectrum management. Our approach uses value functions to evaluate thedesirability of choosing different transmission parameters, and enables efficient assignment of spectrumsand transmit powers by maximizing long-term reward. We then investigate various real-worldscenarios, and compare the communication performance using different sets of learning parameters.The results proves our reinforcement learning based spectrum management can significantly reduceinterference to licensed users, while maintaining a high probability of successful transmissions in acognitive radio ad hoc network.

Dynamic, scalable and flexible resource discovery for large-dimension many-core systems

Future large scale systems will execute novel operating systems running across many chips with many cores. In this highly distributed environment, resource discovery is an important building block. Resource discovery aims to match the application’s demands to the existing (distributed) resources, by discovering and finding resources at run-time, and then selecting the best resource that matches the application running requirements. The main contribution of this paper is the design and evolution of a highly scalable, highly flexible, resource discovery model for such heterogeneous environments. The model is based on self-organizing processing resources in the system according to a hierarchical resource description where each group of resources has a local directory that collects and keeps the information of the underlying resource members (cores) in different layers. Operationally, at each layer, it consists of a peer-to-peer architecture of modules that, by interacting with each other, provide a global view of the resource availability in a large, dynamic and heterogeneous distributed environment. The proposed resource discovery model provides the adaptability and flexibility to perform complex querying by supporting a large set of significant querying features (such as multi-dimensional, range and aggregate querying) while supporting exact and partial matching, both for static and dynamic object contents. The paper demonstrates by simulation how the proposed model can deal with issues such as scalability, efficiency and adaptability of resource discovery in future many-core systems which are the major challenges in the current state of the art. The simulation shows that the proposed resource discovery model can be applied to arbitrary scales of dynamicity, both in terms of complexity and of scale, positioning this proposal as a good architecture for future many-core systems.

ScalaGiST

MapReduce has become the state-of-the-art for data parallel processing. Nevertheless, Hadoop, an open-source equivalent of MapReduce, has been noted to have sub-optimal performance in the database context since it is initially designed to operate on raw data without utilizing any type of indexes. To alleviate the problem, we present ScalaGiST - scalable generalized search tree that can be seamlessly integrated with Hadoop, together with a cost-based data access optimizer for efficient query processing at run-time. ScalaGiST provides extensibility in terms of data and query types, hence is able to support unconventional queries (e.g., multi-dimensional range and k -NN queries) in MapReduce systems, and can be dynamically deployed in large cluster environments for handling big users and data. We have built ScalaGiST and demonstrated that it can be easily instantiated to common B + -tree and R-tree indexes yet for dynamic distributed environments. Our extensive performance study shows that ScalaGiST can provide efficient write and read performance, elastic scaling property, as well as effective support for MapReduce execution of ad-hoc analytic queries. Performance comparisions with recent proposals of specialized distributed index structures, such as SpatialHadoop, Data Mapping, and RT-CAN further confirm its efficiency.

The Design and Implementation of Unified Invoking Component Based on Web Services Framework

Web Services is a platform which enables the applications interoperate on the Internet. It is widely used in designing and building systems in open and dynamic distributed environments such as EAI (Enterprise Application Integration) and B2B (Business to Business). As the development of framework technology, it is convenient and standardized to use framework to develop web applications. For Web Services, the frameworks Axis, Axis2, XFire and CXF are widely used. By the performance testing of the four frameworks, this paper not only introduces the four framework but also analyzes the differences of the four frameworks, and then makes some suggestions for developers to choose the appropriate one. The framework can simplify the development process and decrease the development time. However, because of the differences of the frameworks, the interoperation between different web service’s server and client may cause incompatible problems. This paper analyzes the reasons of this incompatible problems and finally presents a mechanism based on unified invoking component to solve this problems. By parsing and repacking the SOAP messages, the incompatible problems between client and server can be solved successfully.

Distributed Content Semantic Organization and Access Control with EuterpeFS

Files have been the most common information containers for many decades and, as such, they have been established as the natural way for users to access their data. With the growth of distributed systems, the distributed file systems have emerged as a constantly active research area, with various solutions that cover a vast variety of requirements. EuterpeFS is a distributed file system that emphasizes in the efficient semantic organization of the distributed content and its protection, based on MPEG standards. Its layered, fine grained and standards based, design makes EuterpeFS an easily adaptable distributed file system that can work with different publish- subscribe protocols, security and semantic technologies; even with other distributed file systems. The contributions of EuterpeFS can be summarized in the following aspects: ．Semantic organization of content where the directory structured is dictated by ontologies. ．Content protection using MPEG-21 licensing technologies allowing protecting content in a dynamic distributed environment, where users are arbitrarily entering and leaving the system. ．Implementation based on standard MPEG-M API enabling easy integration and interoperation between different implementations of particular components.

Performance optimality enhancement algorithm in DDBS (POEA)

Proper data allocation is a key performance factor for an efficient functionality of Distributed Database Systems (DDBSs). Therefore, if data allocation across sites is performed accurately while preserving the issues of communication and site constraints, an optimal solution for DDBS performance in a dynamic distributed environment will be achieved.In this paper, a new dynamic data allocation algorithm for non-replicated DDBS is presented, the proposed Performance Optimality Enhancement Algorithm (POEA) explores and improves some concepts used in previously developed algorithms to reallocates fragments to different sites given the changing data access patterns, time and sites constraints of the DDBS. Moreover, the POEA adopts the shortest path between the old location and the new anticipated location for the transferred fragments when migration decision is made. Experimental results show that POEA has efficiently reduced the transmission cost subsequently minimizing the frequency and the time spent on fragment migration over the network sites resulting to a great improvement in the overall DDBS performance.

Dynamic service placement and replication framework to enhance service availability using team formation algorithm

The motivation of this work is to reduce the complexity in managing and administering services in the ever growing distributed environment via automated service placement and replication with team formation algorithm. The team formation algorithm is designed in a way that it would continuously search for resources with better performance and pool resources together to achieve better availability. The main intention of this work is not to replace the human administrators but to provide a better alternative in managing services in dynamic distributed environment. Instructions from the administrators are still required but at a different level. Administrators are freed from making low level decisions such as to decide the actual placement of the services and design the failover capabilities for each of the services. The evaluation results showed that the framework is capable of managing resources according to the requirements given by administrator, even during in the event of multiple consecutive resources failure. The proposed solution had the probability of 72.1% of its services that are still available after 83.3% of the available resources were shut down while conventional failover solution using three redundant units had only the probability of 40.8% of services that are still available.

Resource Allocation for Multi-User Cognitive Radio Systems Using Multi-agent Q-Learning

Cognitive Radio (CR) is a new generation of wireless communication system that enables unlicensed users to exploit underutilized licensed spectrum to optimize the radio spectrum utilization. The resource allocation is difficult to achieve in a dynamic distributed environment, in which CR users take decisions to select a channel without negotiation, and react to the environmental changes. This paper focuses on using a multi-agent reinforcement-learning (MARL), Q-learning algorithm, on channels selection decision by secondary users in 2×2 and 3×3 cognitive radio system. Numerical results, obtained with MATLAB, demonstrate that resource allocation is realized without any negotiation between secondary and primary users. In this work, the analogy between the numerical and simulated results is also noted.

Achieving self-managed deployment in a distributed environment

This paper proposes algorithms and mechanisms for achieving self-optimized deployment of computationally intensive scientific and engineering applications in highly dynamic and large-scale distributed environment. The primary focus is on the modeling of the application and underlying architecture into a common abstraction and on the incorporations of autonomic features to those abstractions to achieve self-optimized deployment. To represent the underlying heterogeneous infrastructure, a hierarchical (tree) model of distributed resources has been adopted that organizes distributed nodes in a utility aware way. To accomplish the self-optimization, a utility-function has been formulated that governs both the initial deployment of an application and its dynamic reconfiguration. In our approach, the deployment decisions are made solely based on locally available information and without costly global communication or synchronization. The self-management is therefore decentralized to provide better adaptability, scalability and robustness.

Role-based manufacturing control in a holonic multi-agent system

Multi-agent systems (MAS) are particularly adapted to deal with dynamic distributed environments and are typically used to manage business workflows or data flows in manufacturing systems. From a control perspective, holonic representation allows both the informational and the physical parts in such manufacturing systems to be modelled. In this article, we focus on the notion of role in holonic MAS (HoloMAS) and its contribution to the adaptive control of manufacturing systems. We validated our HoloMAS proposal using simulations and through a real implementation on a flexible assembly cell in our university lab. Our results clearly show that the notion of role helps to adapt and manage perturbations.

Developing Systems that can Utilize Captured Command and Performance Data to Support Collaborative Adaptation in Dynamic Distributed Work Environments

In complex dynamic distributed work environments there is often a need to adapt previously prescribed operations if optimal performance is to be continually achieved. For instance, changes in an organization's personnel, structure, technological resources, procedures, or the physical environment in which work is performed can all have an impact on what adaptation is needed. With modern information systems that capture individual and collaborative decision making and performance in these organizations there is the potential to leverage that data to improve awareness and understanding of what has occurred in the past and to help decide where changes need to be made in future operations. In this paper, two case studies are presented and compared where such data has been utilized this way. The first involves the development of the Post Operations Evaluation Tool to help support adaptation by FAA and Airline operations center staff, and the second involves the development of C2 insight to help support adaptation by Army commanders in command posts. In both cases it was found important to (1) model the system in such a way that contributing factors to performance measures of interest are easily retrieved and appropriately associated with outcome; (2) provide the user with analytical tools to help identify important patterns in the system, but also provide significant freedom in searching through the captured data – starting with meaningful overviews and then providing support for coupled interactive related views on the data, as well as filtering and drill-down capabilities, and (3) provide support for collaborative discussion and problem solving on patterns that one user or group believes are evidence that adaptation is needed.

Metadata for Managing Grid Resources in Data Mining Applications

The Grid is an infrastructure for resource sharing and coordinated use of those resources in dynamic heterogeneous distributed environments. The effective use of a Grid requires the definition of metadata for managing the heterogeneity of involved resources that include computers, data, network facilities, and software tools provided by different organizations. Metadata management becomes a key issue when complex applications, such as data-intensive simulations and data mining applications, are executed on a Grid. This paper discusses metadata models for heterogeneous resource management in Grid-based data mining applications. In particular, it discusses how resources are represented and managed in the Knowledge Grid, a framework for Grid-enabled distributed data mining. The paper illustrates how XML-based metadata is used to describe data mining tools, data sources, mining models, and execution plans, and how metadata is used for the design and execution of distributed knowledge discovery applications on Grids.