Deadlock Prediction Research Articles

Fog computing effective load balancing and strategy for deadlock prediction management

The reliability, real-time performance, and quality of service of fog computing are dramatically affected by deadlock. In this paper, a new Deadlock Management System for a fog computing environment is proposed to achieve the best task scheduling while considering deadlock prediction. It contains five modules; the receiver, the matchmaker, the deadlock predictor, the dispatcher, and the ranking module. The receiver is responsible for incoming tasks, assigning their initial priority and task envelope, and calculating server computational distance. The match-making is a fuzzy inference algorithm used to calculate the server matchmaking degree. This determines the best fog server to perform the task. The deadlock module calculates the deadlock prediction risk degree while The dispatcher module dispatches the task depending on first-fit or best-fit modes. The ranking module is responsible for ranking all suitable fog servers and choosing the most suitable one to implement the task. This paper contributes to providing a new load-balancing strategy in fog computing while taking deadlock into consideration. Thus, resolving drawbacks such as load and complexity that strain the network, and computation latency. The results show that the proposed strategy outperforms other load-balancing techniques reported in the literature.

Sound Dynamic Deadlock Prediction in Linear Time

Deadlocks are one of the most notorious concurrency bugs, and significant research has focused on detecting them efficiently. Dynamic predictive analyses work by observing concurrent executions, and reason about alternative interleavings that can witness concurrency bugs. Such techniques offer scalability and sound bug reports, and have emerged as an effective approach for concurrency bug detection, such as data races. Effective dynamic deadlock prediction, however, has proven a challenging task, as no deadlock predictor currently meets the requirements of soundness, high-precision, and efficiency. In this paper, we first formally establish that this tradeoff is unavoidable, by showing that (a) sound and complete deadlock prediction is intractable, in general, and (b) even the seemingly simpler task of determining the presence of potential deadlocks, which often serve as unsound witnesses for actual predictable deadlocks, is intractable. The main contribution of this work is a new class of predictable deadlocks, called sync(hronization)-preserving deadlocks. Informally, these are deadlocks that can be predicted by reordering the observed execution while preserving the relative order of conflicting critical sections. We present two algorithms for sound deadlock prediction based on this notion. Our first algorithm SPDOffline detects all sync-preserving deadlocks, with running time that is linear per abstract deadlock pattern, a novel notion also introduced in this work. Our second algorithm SPDOnline predicts all sync-preserving deadlocks that involve two threads in a strictly online fashion, runs in overall linear time, and is better suited for a runtime monitoring setting. We implemented both our algorithms and evaluated their ability to perform offline and online deadlock-prediction on a large dataset of standard benchmarks. Our results indicate that our new notion of sync-preserving deadlocks is highly effective, as (i) it can characterize the vast majority of deadlocks and (ii) it can be detected using an online, sound, complete and highly efficient algorithm.

Robust nonlinear model predictive control for automatic train operation based on constraint tightening strategy

AbstractThis paper studies the problem of automatic train operation (ATO) robust nonlinear model predictive control under considering multiple objectives and constraints. After establishing a nonlinear multipoint model with uncertain bounded disturbance, a robust nonlinear model predictive control algorithm to meet the punctuality of train operation and energy consumption for ATO is proposed based on constraint tightening strategy. Moreover, theoretical analysis of the feasibility and stability for the speed loop system are presented. Then, with the objective of reference electromagnetic torque tracking and low switching frequency, a model predictive direct torque control algorithm with one‐step delay compensation is proposed. Feasibility of the proposed algorithm is ensured by using deadlock prediction method, and convergence analysis of the torque loop is given simultaneously. Lastly, the effectiveness of these two algorithms are verified by numerical simulation.

Manipulating deformable objects by interleaving prediction, planning, and control

We present a framework for deformable object manipulation that interleaves planning and control, enabling complex manipulation tasks without relying on high-fidelity modeling or simulation. The key question we address is when should we use planning and when should we use control to achieve the task? Planners are designed to find paths through complex configuration spaces, but for highly underactuated systems, such as deformable objects, achieving a specific configuration is very difficult even with high-fidelity models. Conversely, controllers can be designed to achieve specific configurations, but they can be trapped in undesirable local minima owing to obstacles. Our approach consists of three components: (1) a global motion planner to generate gross motion of the deformable object; (2) a local controller for refinement of the configuration of the deformable object; and (3) a novel deadlock prediction algorithm to determine when to use planning versus control. By separating planning from control we are able to use different representations of the deformable object, reducing overall complexity and enabling efficient computation of motion. We provide a detailed proof of probabilistic completeness for our planner, which is valid despite the fact that our system is underactuated and we do not have a steering function. We then demonstrate that our framework is able to successfully perform several manipulation tasks with rope and cloth in simulation, which cannot be performed using either our controller or planner alone. These experiments suggest that our planner can generate paths efficiently, taking under a second on average to find a feasible path in three out of four scenarios. We also show that our framework is effective on a 16-degree-of-freedom physical robot, where reachability and dual-arm constraints make the planning more difficult.

On Multi-Step Look-Ahead Deadlock Prediction for Automated Manufacturing Systems Based on Petri Nets

Automated manufacturing systems (AMSs) are prone to be in deadlock states when resource allocation is unreasonable. Reasonably allocating system resources to achieve deadlock control is a primary task of the design of an AMS. This paper proposes a multi-step look-ahead deadlock prediction method to obtain an optimal deadlock avoidance policy for a class Petri nets, without calculating a complete reachability graph. As the reachability graph of a large-scale Petri net model is usually large, the analysis and calculation process to obtain the optimal deadlock avoidance policy is complicated. We first simplify a Petri net model by using the existing Petri net reduction methods or removing non-shared resources to simplify the model structure. Then we calculate the dead zone markings of the simplified model through a reverse generation method, and finally develop a multi-step look-ahead deadlock prediction method to obtain an optimal deadlock avoidance policy for a class of AMSs.

Sound deadlock prediction

For a concurrent program, a prediction tool maps the history of a single run to a prediction of bugs in an exponential number of other runs. If all those bugs can occur, then the tool is sound. This is the case for some data race tools like RVPredict, but was, until now, not the case for deadlock tools. We present the first sound tool for predicting deadlocks in Java. Unlike previous work, we use request events and a novel form of executability constraints that enable sound and effective deadlock prediction. We model prediction as a general decision problem, which we show is decidable and can be instantiated to both deadlocks and data races. Our proof of decidability maps the decision problem to an equivalent constraint problem that we solve using an SMT-solver. Our experiments show that our tool finds real deadlocks effectively, including some missed by DeadlockFuzzer, which verifies each deadlock candidate by re-executing the input program. Our experiments also show that our tool can be used to predict more, real data races than RVPredict.

Predicting potential deadlocks in multithreaded programs

SummaryIn a multithreaded program, competition of threads for shared resources raises the deadlock possibility, which narrows the system liveness. Because such errors appear in specific schedules of concurrent executions of threads, runtime verification of threads behavior is a significant concern. In this study, we extended our previous approach for prediction of runtime behavior of threads may lead to an impasse. Such a prediction is of importance because of the nondeterministic manner of competing threads. The prediction process tries to forecast future behavior of threads based on their observed behavior. To this end, we map observed behavior of threads into time‐series data sets and use statistical and artificial intelligence methods for forecasting subsequent members of the sets as future behavior of the threads. The deadlock prediction is carried out based on probing the allocation graph obtained from actual and predicted allocation of resources to threads. In our approach, we use an artificial neural network (ANN) because ANNs enjoy the applicable performance and flexibility in predicting complex behavior. Using three case studies, we contrasted results of the current and our previous approaches to demonstrate results. Copyright © 2015 John Wiley & Sons, Ltd.

Deadlock prevention for automated guided vehicles in automated container terminals

Automated guided vehicles (AGVs) are an important component for automating container terminals. When utilizing AGVs to transport containers from one position to another in a container terminal, deadlocks are a serious problem that must be solved before real operations can take place. This study assumes that the traveling area for AGVs is divided into a large number of grid-blocks, and, as a method of traffic control, grid-blocks are reserved in advance when AGVs are running. The first purpose of the reservation is to make room between AGVs and to prevent deadlocks. The objective of this study is to develop an efficient deadlock prediction and prevention algorithm for AGV systems in automated container terminals. Because the size of an AGV is much larger than the size of a grid-block on a guide path, this study assumes that an AGV may occupy more than one grid-block at a time. This study proposes a method for reserving grid-blocks in advance to prevent deadlocks. A graphical representation method is suggested for a reservation schedule and a priority table is suggested to maintain priority consistency among grid-blocks. It is shown that the priority consistency guarantees deadlock-free reservation schedules for AGVs to cross the same area at the same time. The proposed method was tested in a simulation study.

Cyclic deadlock prediction and avoidance for zone-controlled AGV system

Automated guided vehicles (AGVs) are now becoming popular in automated materials handling systems, flexible manufacturing systems and even containers handling applications at seaports. Its performance affects the efficiency of the entire system. Deadlock formation is a serious problem as it stalls the AGVS. The objective of this paper is to develop an efficient AGV deadlock prediction and avoidance algorithm for container port operations. Deadlock in a broad sense is a situation in which at least a part of the system stalls. There are a lot of situations in which the system may stall and most of these situations can be avoided through the control and navigation system. This paper discusses the development of an efficient strategy for predicting and avoiding the deadlocks in a large scale AGVS A variety of deadlock-detecting algorithms are available, but these methods work mainly for manufacturing system where the network layout is simple and uses only a small number of AGVs. The AGVS under study has complex layout and involves close to 80 AGVs. The proposed solution is implemented using the Automod simulation software, and the performance of the technique is presented. Simulation shows that all the potential deadlock situations can be detected and avoided via this methodology. Also, the proposed strategy is computationally efficient and is able to provide the movement decision to the vehicles within the 1.5-s sampling time.

WORMHOLE DEADLOCK PREDICTION

Deadlock prevention is usually realized by forbidding transmissions that could eventually cause a deadlock to occur. However, an optimal routing algorithm with respect to channel utilization forbids transmissions only when they would bring the network into a configuration that will necessarily evolve into a deadlock. Hence, optimal deadlock prevention is closely related to deadlock prediction. In this paper it is shown that wormhole deadlock and livelock prediction is a hard problem for both oblivious and adaptive routing.

Deadlock prediction and avoidance for zone-control AGVS

Deadlock problems of zone-control uni-directional automated guided vehicle systems (AGVS) are discussed in this paper. A directed graph is defined to model the vehicle flows, to identify distinct vehicle flows, to represent states easily and to capture concurrent behaviour of zone-control AGVS. The motivation of this research was to propose a dynamic approach to enable prediction in real time and to avoid deadlocks that are caused by sharing guidepath zones for a class of AGVS so that both utilization of resource and overall throughput can be improved. The proposed algorithmic procedure, in accordance with the states gained and generated from the obtained model, can serve as a functional module for the operation of a zone-control AGVS without the need to revise the original vehicle control extensively.

Deadlock prediction and avoidance based on Petri nets for zone-control automated guided vehicle systems

Deadlock problems of zone-control uni-directional automated guided vehicle (AGV) systems are discussed in this paper. Deadlocks of two types in such AGV systems are first classified from the perspective of shared resources, i.e. guide-path zones and buffers. A special class of Petri nets, attributed Petri nets (APN), is defined and used to represent the current state and to generate future states of zone-control AGV systems. We propose an algorithmic procedure to predict in real time and to avoid deadlocks that are caused by sharing guide-path zones in zone-control AGV systems. The proposed algorithm utilizes the current system state and future predicted states to avoid deadlocks. These states are obtained and generated from the obtained APN. A modular approach is employed to facilitate the construction of APN models of zone-control AGV systems.

Deadlock prediction for Escrow transactions

The Escrow transactional method permits long-lived transactions to update records without blocking concurrent access by other transactions to records modified. A situation may occur in an Escrow system in which an update request cannot currently be granted, but if the requesting transaction is allowed to wait, the resulting situation is safe in the sense of the banker's algorithm, i.e. a schedule to grant requests exists whereby all current requests can be granted. Any algorithm to find such a schedule must predict potential deadlocks, and we show that this problem is NP-complete, using a variant of a proof by Gold. The proof draws an analogy between transactional deadlock prediction and the banker's algorithm, and thus is of independent interest. In the conclusions of Section 5 we point out some basic negative performance ramifications of deadlock avoidance/detection inherent in the standard approach of the banker's algorithm; this suggests an avenue for further research.

DEADLOCK PREDICTION IN THE CASE OF DYNAMIC ROUTING

One of the main issues in flow control problems is deadlock of messages caused by a limited amount of resources. In this paper, the problem of predicting whether a deadlock will necessarily occur in a Store-and-Forward Network is analyzed. We show that, in the case of dynamic routing, the deadlock prediction problem can be decided in polynomial time if tokens are allowed to transit more than once through the same vertex, in contrast with an NP-completeness result in the case where they are allowed to transit at most once.

Deadlock Prediction: Easy and Difficult Cases

The subject of this paper is the computational complexity of the deadlock prediction problem for resource allocation. This problem is the question “Is deadlock avoidable?” i.e. “Is there a feasible sequence in which to allocate all the resource requests?” given the current status of a resource allocation system. This status is defined by (1) the resource vector held by the banker, i.e. the quantity of resources presently available for allocation, and (2) the resource requests of the processes: Each process is required to make a termination request of the form “Give me resource vector y and I will eventually terminate and return resource vector z.” Also, each process can make any number of partial requests of the form “If you can’t give me y, then give me a smaller resource vector $y'$ and I will be able to reach a point at which I can halt and temporarily return to $z'$, although I will still need need $y - y' + z'$ to terminate.” If (1) the resources are reusable and (2) partial requests are not allowed, then Habermann’s well known “banker’s algorithm” solves the deadlock prediction problem in polynomial time. The results of this paper are as follows: If either constraint (1) or constraint (2) is removed, then the deadlock prediction problem is NP-complete. A generalization of the banker’s algorithm solves this problem with the following constraint: For each process, the 0 vector and the profitability vectors $(z - y)$ of its partial and termination requests are linearly ordered with respect to size. This case includes systems with one type of resource or with Hansen’s (1973) hierarchically partitioned reusable resources.