As the adoption of microservices architectures in cloud deployments grows, so does the challenge of ensuring rapid recovery and minimal energy consumption in the face of disasters. This paper introduces a Generalized Stochastic Petri Net (GSPN) model specifically designed to quantify recovery times and electrical consumption in such environments. The model supports the development of resilient and eco-conscious systems by allowing precise manipulation and planning based on various configuration scenarios. We identify critical architectural elements and define intervals that yield significant improvements. Our findings not only enhance the understanding of energy and recovery dynamics in microservices but also serve as a crucial tool for system designers aiming to optimize both performance and sustainability. The implications of this research facilitate a strategic approach to disaster recovery planning, contributing to the broader field of cloud computing resilience.

This study investigates the loss of generalization ability in neural networks, revisiting warm-starting experiments from Ash & Adams. Our empirical analysis reveals that common methods designed to enhance plasticity by maintaining trainability provide limited benefits to generalization. While reinitializing the network can be effective, it also risks losing valuable prior knowledge. To this end, we introduce the Hare & Tortoise, inspired by the brain's complementary learning system. Hare & Tortoise consists of two components: the Hare network, which rapidly adapts to new information analogously to the hippocampus, and the Tortoise network, which gradually integrates knowledge akin to the neocortex. By periodically reinitializing the Hare network to the Tortoise's weights, our method preserves plasticity while retaining general knowledge. Hare & Tortoise can effectively maintain the network's ability to generalize, which improves advanced reinforcement learning algorithms on the Atari-100k benchmark. The code is available at https://github.com/dojeon-ai/hare-tortoise.

Hyperledger Fabric stands as a leading framework for permissioned block-chain systems, ensuring data security and audit-ability for enterprise applications. As applications on this platform grow, understanding its complex configuration concerning various block-chain parameters becomes vital. These configurations significantly affect the system’s performance and cost. In this research, we introduce a Stochastic Petri Net (SPN) model to analyze Hyper-ledger Fabric’s performance, considering variations in block-chain parameters, computational resources, and transaction rates. We provide case studies to validate the utility of our model, aiding block-chain administrators in determining optimal configurations for their applications. A key observation from our model highlights the block size’s role in system response time. We noted an increased mean response time, between 1 to 25 seconds, due to variations in transaction arrival rates.

Home automation or domotics is a typical representative of everything as a service (XaaS). Individual houses are equipped with Internet of Things (IoT) sensors and home facilities capable of self-assessment to offer comfortable, secure, and high-quality home services to their residents. However, assessing such systems with a high level of diversity is paramount of importance and challenging to assimilate. Domotics XaaS requires a high quality of service (QoS) in service performance and operational availability. In that regard, we propose, in this paper, a modeling approach based on stochastic Petri nets (SPN) for the performability quantification of domotics architectures. SPN performability models are developed following the architecture of a home automation system consisting of several IoT sensors/devices to evaluate the trade-offs between performance and availability of home automation services. The inter-dependency between performance and availability metrics is evaluated. The metrics include, for example, the mean response time (MRT) and the number of discarded packets. Sensitivity analysis using the design of experiments (DoE) is performed to identify the system's impacting components and performability bottleneck. Simulation results highlight the useful aspects of the proposed performability models for architectural and operational optimization of home automation XaaS infrastructures.

Rising aggressive virus pandemics urge to conduct studies on dependability and security of modern computing systems to secure autonomous and continuous operations of healthcare systems. In that regard, we propose to quantify dependability and security measures of an Internet-of-Medical Things (IoMT) infrastructure relied on an integrated physical architecture of cloud/fog/edge (CFE) computing paradigms in this article. We propose a reliability/availability quantification methodology for the IoMT infrastructure using a hierarchical model of three levels: 1) fault tree (FT) of overall IoMT infrastructure consisting of CFE member systems; 2) FT of subsystems within CFE member systems; and 3) continuous-time Markov chain (CTMC) models of components/devices in the subsystems. We incorporate a number of failure modes for the underlying subsystems, including Mandel-bug related failures and non-Mandel bugs related failure, as well as failures due to cyber-security attacks on software subsystems. Five case-studies of configuration alternation and four operational scenarios of the IoMT infrastructure are considered to comprehend the dependability characteristics of the IoMT physical infrastructure. The metrics of interest include reliability over time, steady state availability (SSA), sensitivity of SSA wrt. selected mean time to failure—equivalent (MTTFeq) and mean time to recovery—equivalent (MTTReq), and sensitivity of SSA wrt. frequencies of cyber-security attacks on software subsystems. The analysis results help comprehend operational behaviors and properties of a typical IoMT infrastructure. The findings of this study can improve the design and implementation of real-world IoMT infrastructures consisting of cloud, fog, and edge computing paradigms.

In Robotics field communication is served as one of the key objectives. Communication plays significant role between controllers and robots, and during collaboration of number of robots. Providing safe and reliable communication between elements of collaborative environment had become important issue. The UCR Reconfigurable Communication Broker has been presented as a major contribution of this research work. This solution contains all functionalities of previous It was designed based on OSGi platform and all additional components can be built as a standard OSGi bundle. It makes system more flexible to run its bundles in a real-time. As a final instance result of this work Dynamic Reconfigurable Lightweight Broker was developed. It can run on different platforms, as an example it is demonstrated on Android platform, to control sort of robots by installing different bundles.

RUAV (Rotorcraft based Unmanned Aerial Vehicle) have been exploited in various fields such as surveillance, reconnaissance and search. Most of the study in this area is focused on single RUAV; however using multiple unmanned vehicles is big advantage to accomplish the mission in a short time and effective way. Moreover little amount of research has been undertaken in the development of multiple RUAV control systems. This paper proposes a behavior-based decentralized approach that allows an RUAV to carry out its own mission of flying to a specified region while the distances between RUAVs are maintained constantly to avoid collision. The main goal of the proposed controller is to make the RUAV cooperate among each other to achieve the defined task. In this research Reynold's flocking model based behavior approach has been utilized. Along with this the testing environment for multiple RUAV is developed to validate proposed control algorithm. Complete setup is implemented and run under QNX RTOS, based on PC104 embedded board. The multiple RUAV is tested and evaluated using HILS (Hardware-In-the-Loop Simulation). To validate the proposed approach, simulation is performed to achieve the waypoint with multiple RUAVs.

Abstract The prosperity of Internet results in a workflow executing engine's performance instability due to the request congestion for a short period of time. This paper proposes a peak load control (PLC) based orchestration system that can stably execute hybrid services. The PLC mechanism uses the delay time algorithm for controlling a BPEL engine's heavy peak load caused by the request congestion for a short period of time. In order to prove the stable performance of the PLC‐based orchestration system, we analyze the proposed delay time algorithm. According to our experimental results, the proposed delay time algorithm can stably execute structured activities of WSBPEL specification in heavily overloaded state after the saturation phase and has an effect on controlling the states of peak load. In this paper, we also describe a hybrid service architecture model that can represent both Web Services and existing EAs as same type of services. Copyright © 2007 John Wiley & Sons, Ltd.

In this paper, we present an application of self-healing systems in the form of service request optimization algorithm for service gateways. This algorithm preestimates the load of the gateway and accepts the service request according to the capability of the worker thread. The experimental and simulation results show that the performance of load control algorithm works well for managing the service request. Our study has revealed that normally the networks fail to entertain the requests after the saturation point. Whereas the algorithm proposed in this paper, virtually accepts the service requests almost always which is what is essentially required from hybrid network in a ubiquitous environment. The algorithm proposed is already embedded and are in work in the IT infrastructure of major banks of Korea.

The paper describes the design and implementation of Real-Time Inter-ORB Protocol suited for multimedia applications on the CORBA environment. Conventional CORBA is not well suited for real time applications, since it does not define policies nor mechanisms to provide QoS guarantees, and does not define key features that are necessary to support real time programming. The existing TAO's RIOP for Real-Time CORBA has a mechanism to transfer QoS information but has a lack of support for QoS enforcement to achieve those QoS requests. In order to solve this problem, the paper suggests an integration of GIOP/IIOP with RSVP for Real-Time CORBA. A key feature is presented: QoS enforcement, which is provided for Real-Time CORBA by integrating IIOP with RSVP.