Self-Scheduling for a Heterogeneous Distributed Platform

With the rapid development of new power systems, diverse new power services have imposed stricter requirements on network resources and performance. However, the traditional method of transmitting request data to the IoT management platform for unified processing suffers from large delays due to long transmission distances, making it difficult to meet the delay requirements of new power services. Therefore, to reduce the transmission delay, data transmission, storage and computation need to be performed locally. However, due to the limited resources of individual nodes in the local power wireless communication network, issues such as tight coupling between devices and resources and a lack of flexible allocation need to be addressed. The collaborative allocation of resources among multiple nodes in the local network is necessary to satisfy the multi-dimensional resource requirements of new power services. In response to the problems of limited node resources, inflexible resource allocation, and the high complexity of multi-dimensional resource allocation in local power wireless communication networks, this paper proposes a multi-objective joint optimization model for the collaborative allocation of communication, storage, and computing resources. This model utilizes the computational characteristics of communication resources to reduce the dimensionality of the objective function. Furthermore, a mouse swarm optimization algorithm based on multi-strategy improvements is proposed. The simulation results demonstrate that this method can effectively reduce the total system delay and improve the utilization of network resources.

In order to cope with the ever-increasing traffic demands and stringent latency constraints, next generation, i.e., sixth generation (6G) networks, are expected to leverage Network Function Virtualization (NFV) as an enabler for enhanced network flexibility. In such a setup, in addition to the traditional problems of user association and traffic routing, Virtual Network Function (VNF) placement needs to be jointly considered. To that end, in this paper, we focus on the joint network and computational resource allocation, targeting low network power consumption while satisfying the Service Function Chain (SFC), throughput, and delay requirements. Unlike the State-of-the-Art (SoA), we also take into account the Access Network (AN), while formulating the problem as a general Mixed Integer Linear Program (MILP). Due to the high complexity of the proposed optimal solution, we also propose a low-complexity energy-efficient resource allocation algorithm, which was shown to significantly outperform the SoA, by achieving up to 78% of the optimal energy efficiency with up to 742 times lower complexity. Finally, we describe an Orchestration Framework for the automated orchestration of vertical-driven services in Network Slices and describe how it encompasses the proposed algorithm towards optimized provisioning of heterogeneous computation and network resources across multiple network segments.

The computing process of tasks in Industrial Internet of Things (IIoT) environments is becoming increasingly complex due to the development of 5G and artificial intelligence. Leading devices are increasingly relying on heterogeneous platforms that integrate different types of processing units, such as CPUs, GPUs, and other resources, to meet the requirements of delay-sensitive and computing-intensive tasks. However, compared to conventional general-proposed CPU computing, CPU–GPU heterogeneous computing typically involves three processes, i.e., task preprocessing, hybrid computing, and result aggregation. These processes are associated with particular computing resources, which increases the difficulty of task offloading and computing resource allocation under task-specific resource and delay constraints. In this article, we first propose a three-stage heterogeneous computing (TSHC) model to practically describe the computing process of parallelizable tasks. Considering the heterogeneous computing resources, device queue backlogs, and collaboration of multiple edge servers, the joint task offloading and heterogeneous resource allocation (JCOHRA) problem is formulated to minimize the long-term average delay of tasks. Then, the Lyapunov optimization method is adopted to simplify the long-term queue stability constraint to a single-slot dynamic optimization problem, which is then modeled as a Markov decision process (MDP). Owing to the tight coupling between decision variables and enormous action space, we propose the multihead proximal policy optimization (MH-PPO)-based JCOHRA algorithm, which is enabled by elaborate constraint transformation and reward function design. Simulation results demonstrate that the JCOHRA scheme achieves better performance than baseline methods in minimizing the long-term average delay of tasks.

Today's typical computational environment consists of a large numbers of commodity workstations interconnected by high bandwidth networks. However, only a small portion of workstation's capacity is utilized since a large share of workstations may be idle at any given moment. Moreover, the communication subsystem dedicated to the idle workstations may be also unused. This results in immense aggregate waste of computational and network resources. We propose the general design of a runtime that executes a standard Java application on available idle workstations. This runtime transparently distributes the application's threads and objects among the currently idle nodes.Our target environment contains a very large but fluctuating number of non-dedicated workstations. The biggest obstacle in such a non-dedicated environment is a participating workstation becoming unable to continue its part of the execution. This can be caused by a failure, e.g., a power outage, etc., or by user intervention, e.g., shutdown or restart. These events result in the immediate loss of the data and threads located on the node, which violates the integrity of the distributed execution. Another, at first seemingly less severe problem is minimizing the response time impact when a node is reclaimed by its owner. This is important, because if the response time impact is not negligible the user will forbid participation of his/her workstation in the distributed executio.The described system possesses the following features. First, it is fault tolerant, preserving the correctness of the distributed execution even if a large portion of nodes fails. By use of the system's fault tolerance we make it non-intrusive with respect to the owners of the participating nodes. We reduce impact on the node's response time to essentially zero by modeling the node's reclamation as a failure, killing all runtime threads and freeing runtime memory as soon as presence of node's owner is detected.Second, our runtime is completely transparent to the programmer and can execute a standard, pre-existing, multithreaded Java program written for a single machine This eliminates the need for a Java programmer to study a new programming paradigm or to use unfamiliar libraries in order to deal with the true face of a non-dedicated distributed environment.Third, the proposed runtime is orthogonal to the implementation of the JVM. It does not require any changes to a node's JVM. This is achieved by rewriting the application's bytecodes, tying it to the bytecodes of the runtime. The instrumented program becomes a distributed application, ready to be executed on multiple nodes using only the node's standard JVM. The ability to use standard JVMs allows portability across heterogeneous collections of commodity workstations. Moreover, it allows each node to locally optimize the performance of its JVM, e.g., via a JIT.Finally, the described runtime is highly scalable, using an efficient and scalable multithreaded software distributed shared memory (DSM) and a fault tolerance scheme that does not require global cooperation. Scalability allows our runtime to incorporate a large number of workstations increasing its computational power.We view this work as a first step in providing a convenient computing infrastructure using wide-scale non-dedicated environments. The underlying question is whether (mostly idle) Internet and enterprise interconnects are sufficiently wide to efficiently support high-level programming paradigms, such as distributed shared memory. If this is the case, then, as opposed to current community wisdom, it is not mandatory to restrict the communication pattern of a program that is utilizing idle resources. Java, as a popular multithreaded programming language, is best suited for this experiment.

An allocation of resources to a virtual machine specifies the maximum amount of each individual element of each resource type that will be utilized, as well as the aggregate amount of each resource of each type. An allocation is thus represented by two vectors, a maximum elementary allocation vector and an aggregate allocation vector. There are more general types of resource allocation problems than those we consider here. In this paper, we present an approach for improving parallel deadlock detection algorithm, to schedule the policies of resource which supply for resource allocation in heterogeneous distributed platform. Parallel deadlock detection algorithm has a run time complexity of O(min(m,n)), where m is the number of resources and n is the number of processes. We propose the algorithm for allocating multiple resources to competing services running in virtual machines on a heterogeneous distributed platform. The experiments also compare the performance of the proposed approach with other related work.

Static scheduling of multiple workflows with soft deadlines in non-dedicated heterogeneous environments

Computer-aided drug design (CADD) has revolutionized the drug discovery arena and it has reduced the costs associated with finding novel compounds which are having pharmaceutical importance. In CADD, the scientists use the computer software to discover biological active compounds. Molecular docking and energy minimization tools are essential components of structure based drug design. It is a significant tool in structural molecular biology and computer-assisted drug design. It reduces the laboratory workload of the end user and allows researchers to restrict their docking studies to the smallest and the most representative set of macromolecules and small molecules possible. This greatly enhances the productivity of researchers. Energy minimization is an important criterion for selecting a potential 3D molecule. In modeled structures, the 3D structure is affected is due to steric clashes. These clashes happen in a protein structure due to the overlap of non bonding atoms and with the assistance of energy minimization, steric clashes can be eradicated. The open software’s and databases provides a platform for scientists and scholars to carry out their research work in a better way. The docking tools are discussed in this review cover protein-ligand, protein-peptide as well as protein-nucleic acid docking. The tools described include AutoDock 4 and Vina, UCSF DOCK, FLIPDock, EADock, HADDOCK 2.2, SwissDock, PatchDock and ClusPro. In addition to the docking tools, energy minimization tools such as YASARA minimization server, KoBaMIN server and 3D refine server have also been discussed. This mini-review concentrates on open software tools which are free of cost and can be easily downloaded in the computers that are useful for CADD.
 Keywords: Molecular docking; Energy minimization; Structure refinement; Drug design; CADD

Heterogeneity has emerged as one of the most profound and challenging characteristics of today's parallel environments. From the macro level, where networks of distributed computers, composed of diverse node architectures, are interconnected with potentially heterogeneous networks, to the micro level, where deeper memory hierarchies and various accelerator architectures are increasingly common, the impact of heterogeneity on all computing tasks is increasing rapidly. Traditional parallel algorithms, programming environments, and tools, designed for legacy homogeneous multiprocessors, achieve a small fraction of the efficiency and the potential performance that can be obtained in current and future heterogeneous computing platforms. New ideas, innovative algorithms, and specialized programming environments and tools are needed to efficiently use these modern parallel and heterogeneous architectures. The International workshop on algorithms, models and tools for parallel computing on heterogeneous platforms (HeteroPar) is a forum for researchers working on algorithms, programming languages, tools, and theoretical models for efficiently solving complex problems on heterogeneous parallel platforms. HeteroPar 2021 took place (virtually) in Lisbon, Portugal, organized for the 13th time in conjunction with the Euro-Par annual international conference. The format of the workshop included one keynote and 11 technical presentations. The workshop papers represented an interesting mix of topics, addressing the implementation of algorithms and kernels for heterogeneous computing, programming models, data management, runtime and resource management, energy efficiency, cloud computing, and artificial intelligence-based methods oriented towards heterogeneous platforms, as the basis for the next generation exascale computers. The selected papers cover a good spectrum of research topics in the area of heterogeneous computing showing the challenges present on these modern platforms.

We present an innovative path following system based upon multi-camera visual odometry and visual landmark matching. This technology enables reliable mobile robot navigation in real world scenarios including GPS-denied environments both indoors and outdoors. We recover paths in full 3D, making it applicable to both on and off-road ground vehicles. Our controller relies on pose updates from visual odometry, allowing us to achieve path following even when only a joystick drive interface to the base robot platform is available. We experimentally investigate two specific applications of our technology to autonomous navigation on ground vehicles - non line-of-sight leader-following (between heterogeneous platforms) and retro-traverse to home base. For safety and reliability we add dynamic short range obstacle detection and reactive avoidance capabilities to our controller. We show the results for end-to-end real time implementation of this technology using current off-the-shelf computing and network resources in challenging environments.

Scheduling real-time applications modelled as directed acyclic graphs on heterogeneous distributed platforms is known to be a challenging as well as a computationally demanding problem. This article deals with the design of an efficient scheduler for executing a real-time task graph on a distributed platform consisting of a set of fully connected heterogeneous processors. The objective of the scheduling strategy is to minimize a <i>generic penalty function</i> which can be amicably adopted toward its deployment in various application domains such as real-time embedded systems, cloud/fog computing, industrial automation and IoTs, smart grids, automotive and avionic systems, etc. We have first encoded the problem as a constraint satisfaction problem and then developed an efficient list-based heuristic scheduling algorithm called <i>Penalty-aware REal-time Scheduler for Task graphs on heterOgeneous platforms</i> (<i>PRESTO</i>), to generate a minimal penalty deadline-meeting static schedule. The generic efficacy of <i>PRESTO</i> is exhibited through extensive simulation-based experiments using standard benchmark task graphs. The practical applicability of <i>PRESTO</i> in diverse scenarios have further been exhibited by using the scheme in two different real-world case studies, the first of which relates to automotive embedded systems, while the second is in the domain of fog computing.

Introduction/background summary: Communities of practice (CoPs) are groups of practitioners who share a domain of interest and strive to learn from each other by interacting regularly in order to improve practice. CoP as a methodology can reveal, clarify and support the development of new roles and innovative practices rooted in local contexts and networks and contribute to more integration of social and health organizations. Targeted population: In this case, the target population of CoPs is practitioners, organizations and communities willing to develop innovative practices and roles for better integration of health and social care for patients/users. Participants: In Brussels, four CoPs –each anchored in a different district- bring together twelve practitioners from different social and health organizations and disciplines, and two moderators. A total of forty-eight practitioners and eight moderators participate in the CoPs. Description of the initiative: Members of the CoP are enrolled in monthly meetings throughout the year where they share difficulties, resources and find solutions together, creating value through social learning. These health, care, and community workers identify as ‘Trusted frontline professionals”. In this role, practitioners aim to support and guide the patient/user according to his/her needs and goals through the social and health system. CoPs are guided by moderators who use different methods and tools to facilitate exchanges and to support CoP members in their ‘Trusted frontline professional’ role. Thematic analysis is in progress using NVIVO 12 on data collected during CoP and moderators’ meetings through participatory observation, as well as on CoP monitoring tools, conventions and previous reports. A documentation grid was developed through a participatory process within the CoPs. Results: Data from thirty-five CoP meetings, involving forty-eight practitioners will provide insights categorized by established areas of the documentation grid: 1) profile of patients/users supported and practitioners involved; 2) types of situations encountered and actions; 3) proposed solutions and clues from other CoP members; 4) available tools and resources in local networks; 5) development of professional stance; and 6) impact of participation on CoP members and represented organizations. Data from seven moderators’ meetings highlight key elements of CoP organization and facilitation methods, which will be compiled in a methodological guide. Learning for the international audience: CoP as a valuable methodology for exploring and defining innovative roles such as that of ‘Trusted frontline professional’. Social learning processes reveal specific needs of patients/users and practitioners playing this role, resources and tools available to practitioners. Social learning creates value by helping practitioners to be better equipped to play the role of ‘Trusted frontline professional' within local contexts and networks. Next steps: Capturing and capitalising on the value created within these four CoPs can be used to build further CoPs to support practitioners in the 'trusted frontline professional' role in other local areas or to develop and monitor the implementation of other roles.

In this paper, an algorithm for planning and distribution heterogeneous resources in a stochastic system with ordered sets of tasks is proposed. Processing resources and memory resources are considered. The algorithm is based on the GERT-network model with a new parameter – a random variable of resource. A system consisting of many parallel processes is described in the model. Each process includes multiple tasks. By finding the start and end times of a pair of tasks it is possible to determine whether one task can start when another has finished. If the probability is satisfied for a given value, then that pair of tasks can be performed by the same processing resource unit. The ordered set of multiple sequential tasks is the desired outcome. Each set comes with a probability of completing all the tasks in the set with only one processing resource unit. The fewer the number of sets, the more resources are saved. However, it is necessary to consider the probability of the set. Accepting a certain level of risk can save a corresponding number of resources. To support the computation process, an algorithm to find the distribution laws of any arc in the GERT-network has been introduced. It speeds up the computation and allows selected arc located on the loop. Assuming that the memory resources are released as soon as the task is complete, the total memory resources for the system are recalculated. An example of the analysis process, planning and distribution resources in network devices is given.

Chapter twelve - Networking

Crimean–Congo hemorrhagic fever virus (CCHFV), the fatal human pathogen is transmitted to humans by tick bite, or exposure to infected blood or tissues of infected livestock. The CCHFV genome consists of three RNA segments namely, S, M, and L. The unusual large viral L protein has an ovarian tumor (OTU) protease domain located in the N terminus. It is likely that the protein may be autoproteolytically cleaved to generate the active virus L polymerase with additional functions. Identification of the epitope regions of the virus is important for the diagnosis, phylogeny studies, and drug discovery. Early diagnosis and treatment of CCHF infection is critical to the survival of patients and the control of the disease. In this study, we undertook different in silico approaches using molecular docking and immunoinformatics tools to predict epitopes which can be helpful for vaccine designing. Small molecule ligands against OTU domain and protein–protein interaction between a viral and a host protein have been studied using docking tools.

A real-time task scheduling system model was analyzed under a heterogeneous multiprocessor platform with task duplication. This analysis focused on the designs and performances of linear and dynamic programming algorithms for real-time task scheduling under a heterogeneous platform with task duplication. Moreover, experimental analyses were performed to evaluate the performances of different algorithms under different conditions. The advantages of the two proposed algorithms were compared under the same situations to discover which one achieves a higher task scheduling efficiency for a heterogeneous real-time system.