Distributed algorithm

About: Distributed algorithm is a research topic. Over the lifetime, 20416 publications have been published within this topic receiving 548109 citations.

Subgradient methods and consensus algorithms for solving convex optimization problems

Abstract: In this paper we propose a subgradient method for solving coupled optimization problems in a distributed way given restrictions on the communication topology. The iterative procedure maintains local variables at each node and relies on local subgradient updates in combination with a consensus process. The local subgradient steps are applied simultaneously as opposed to the standard sequential or cyclic procedure. We study convergence properties of the proposed scheme using results from consensus theory and approximate subgradient methods. The framework is illustrated on an optimal distributed finite-time rendezvous problem.

Distributed Cooperative Attitude Synchronization and Tracking for Multiple Rigid Bodies

Abstract: In this paper, distributed cooperative attitude synchronization and tracking problems are considered for multiple rigid bodies with attitudes represented by modified rodriguez parameters. Two distributed control laws are proposed and analyzed. The first control law applies a passivity approach for distributed attitude synchronization. The control law guarantees attitude synchronization without the requirement for absolute angular velocity measurements and relative angular velocity measurements between neighboring rigid bodies. The second control law incorporates a time-varying reference attitude, where the reference attitude is allowed to be available to only a subset of the group members under general directed information exchange. The control law guarantees that all rigid bodies track the time-varying reference attitude as long as a virtual leader whose attitude is the time-varying reference attitude has a directed path to all other rigid bodies in the group. Simulation results are presented to demonstrate the effectiveness of the two control laws.

Optimal Scaling of a Gradient Method for Distributed Resource Allocation

Abstract: We consider a class of weighted gradient methods for distributed resource allocation over a network. Each node of the network is associated with a local variable and a convex cost function; the sum of the variables (resources) across the network is fixed. Starting with a feasible allocation, each node updates its local variable in proportion to the differences between the marginal costs of itself and its neighbors. We focus on how to choose the proportional weights on the edges (scaling factors for the gradient method) to make this distributed algorithm converge and on how to make the convergence as fast as possible. We give sufficient conditions on the edge weights for the algorithm to converge monotonically to the optimal solution; these conditions have the form of a linear matrix inequality. We give some simple, explicit methods to choose the weights that satisfy these conditions. We derive a guaranteed convergence rate for the algorithm and find the weights that minimize this rate by solving a semidefinite program. Finally, we extend the main results to problems with general equality constraints and problems with block separable objective function.

Introduction to Reliable and Secure Distributed Programming

Abstract: In modern computing a program is usually distributed among several processes. The fundamental challenge when developing reliable and secure distributed programs is to support the cooperation of processes required to execute a common task, even when some of these processes fail. Failures may range from crashes to adversarial attacks by malicious processes.Cachin, Guerraoui, and Rodrigues present an introductory description of fundamental distributed programming abstractions together with algorithms to implement them in distributed systems, where processes are subject to crashes and malicious attacks. The authors follow an incremental approach by first introducing basic abstractions in simple distributed environments, before moving to more sophisticated abstractions and more challenging environments. Each core chapter is devoted to one topic, covering reliable broadcast, shared memory, consensus, and extensions of consensus. For every topic, many exercises and their solutions enhance the understanding This book represents the second edition of "Introduction to Reliable Distributed Programming". Its scope has been extended to include security against malicious actions by non-cooperating processes. This important domain has become widely known under the name "Byzantine fault-tolerance".

Maximizing throughput in wireless networks via gossiping

Abstract: A major challenge in the design of wireless networks is the need for distributed scheduling algorithms that will efficiently share the common spectrum. Recently, a few distributed algorithms for networks in which a node can converse with at most a single neighbor at a time have been presented. These algorithms guarantee 50% of the maximum possible throughput. We present the first distributed scheduling framework that guarantees maximum throughput. It is based on a combination of a distributed matching algorithm and an algorithm that compares and merges successive matching solutions. The comparison can be done by a deterministic algorithm or by randomized gossip algorithms. In the latter case, the comparison may be inaccurate. Yet, we show that if the matching and gossip algorithms satisfy simple conditions related to their performance and to the inaccuracy of the comparison (respectively), the framework attains the desired throughput.It is shown that the complexities of our algorithms, that achieve nearly 100% throughput, are comparable to those of the algorithms that achieve 50% throughput. Finally, we discuss extensions to general interference models. Even for such models, the framework provides a simple distributed throughput optimal algorithm.