Distributed Constraint Optimization Algorithms Research Articles

Resilient Distributed Constraint Reasoning to Autonomously Configure and Adapt IoT Environments

In this article, we investigate multi-agent techniques to install autonomy and adaptation in IoT-based smart environment settings, like smart home scenarios. We particularly make use of the smart environment configuration problem (SECP) framework, and map it to a distributed optimization problem (DCOP). This consists in enabling smart objects to coordinate and self-configure as to meet both user-defined requirements and energy efficiency, by operating a distributed constraint reasoning process over a computation graph. As to cope with the dynamics of the environment and infrastructure (e.g., by adding or removing devices), we also specify the k -resilient distribution of graph-structured computations supporting agent decisions, over dynamic and physical multi-agent systems. We implement a self-organizing distributed repair method, based on a distributed constraint optimization algorithm to adapt the distribution as to ensure the system still performs collective decisions and remains resilient to upcoming changes. We provide a full stack of mechanisms to install resilience in operating stateless DCOP solution methods, which results in a robust approach using a fast DCOP algorithm to repair any stateless DCOP solution methods at runtime. We experimentally evaluate the performances of these techniques when operating stateless DCOP algorithms to solve SECP instances.

A distributed constraint multi-agent model for water and reclaimed wastewater allocation in urban areas: Application of a modified ADOPT algorithm.

Distributed Constraint Optimization (DCOP)-based approaches, as the distributed version of constraint optimization, provide a framework for coordinated decision making by a team of agents. In this paper, an agent-based DCOP model is developed to allocate water and reclaimed wastewater to demands considering the conflicting interests of involved stakeholders. One of the well-known DCOP algorithms, ADOPT1, is modified to incorporate an agent responsible for monitoring and conserving water resources. This new algorithm considers the social characteristics of agents and a new form of interaction between agents. For the first time in the literature, a real-world water and reclaimed wastewater allocation problem is formulated as a DCOP and solved using the Modified ADOPT (MADOPT) algorithm. To evaluate the MADOPT algorithm, it is applied to a water and reclaimed wastewater allocation problem in Tehran, Iran. The results illustrate the applicability and efficiency of the proposed methodology in dealing with large-scale multi-agent water resources systems. It is also shown that agents' selfishness and social relationships could affect their water use policies.

A local stability supported parallel distributed constraint optimization algorithm.

This paper presents a new distributed constraint optimization algorithm called LSPA, which can be used to solve large scale distributed constraint optimization problem (DCOP). Different from the access of local information in the existing algorithms, a new criterion called local stability is defined and used to evaluate which is the next agent whose value needs to be changed. The propose of local stability opens a new research direction of refining initial solution by finding key agents which can seriously effect global solution once they modify assignments. In addition, the construction of initial solution could be received more quickly without repeated assignment and conflict. In order to execute parallel search, LSPA finds final solution by constantly computing local stability of compatible agents. Experimental evaluation shows that LSPA outperforms some of the state-of-the-art incomplete distributed constraint optimization algorithms, guaranteeing better solutions received within ideal time.

Dynamic multiagent load balancing using distributed constraint optimization techniques

Resource management is a key challenge in multiagent systems. It is especially important in dynamic environments where decisions need to be made quickly and when decisions can get obsolete quickly. In wireless local area networks WLANs, resource management includes dynamic channel assignment, dynamic transmit power control and load balancing of WLANs traffic. In this work, we present a novel decentralized framework and a distributed optimization algorithm DLB-SDPOP for load balancing in complex WLANs. Our self-stabilizing algorithm focuses on repairing the network structure instead of reconstructing it when network perturbations occur. It controls the complexity of problem solving by utilizing efficient search and leverages uncertainty to reduce the possibility of reaching myopic solutions. The size and number of inter-agent communication messages are significantly reduced using a communication filtering mechanism. We categorize different scenarios based on key characteristics i.e., load, dynamics and uncertainty in WLANs and compare our algorithm with other state of the art distributed constraint optimization algorithms in each scenario. Our empirical results show that our distributed approach improves WLANs load balancing performance significantly.

DISTRIBUTED ON-LINE MULTI-AGENT OPTIMIZATION UNDER UNCERTAINTY: BALANCING EXPLORATION AND EXPLOITATION

A significant body of work exists on effectively allowing multiple agents to coordinate to achieve a shared goal. In particular, a growing body of work in the Distributed Constraint Optimization (DCOP) framework enables such coordination with different amounts of teamwork. Such algorithms can implicitly or explicitly trade-off improved solution quality with increased communication and computation requirements. However, the DCOP framework is limited to planning problems; DCOP agents must have complete and accurate knowledge about the reward function at plan time. We extend the DCOP framework, defining the Distributed Coordination of Exploration and Exploitation (DCEE) problem class to address real-world problems, such as ad-hoc wireless network optimization, via multiple novel algorithms. DCEE algorithms differ from DCOP algorithms in that they (1) are limited to a finite number of actions in a single trial, (2) attempt to maximize the on-line, rather than final, reward, (3) are unable to exhaustively explore all possible actions, and (4) may have knowledge about the distribution of rewards in the environment, but not the rewards themselves. Thus, a DCEE problem is not a type of planning problem, as DCEE algorithms must carefully balance and coordinate multiple agents' exploration and exploitation. Two classes of algorithms are introduced: static estimation algorithms perform simple calculations that allow agents to either stay or explore, and balanced exploration algorithms use knowledge about the distribution of the rewards and the time remaining in an experiment to decide whether to stay, explore, or (in some algorithms) backtrack to a previous location. These two classes of DCEE algorithms are compared in simulation and on physical robots in a complex mobile ad-hoc wireless network setting. Contrary to our expectations, we found that increasing teamwork in DCEE algorithms may lower team performance. In contrast, agents running DCOP algorithms improve their reward as teamwork increases. We term this previously unknown phenomenon the team uncertainty penalty, analyze it in both simulation and on robots, and present techniques to ameliorate the penalty.

Cluster Tree Elimination for Distributed Constraint Optimization with Quality Guarantees

Some distributed constraint optimization algorithms use a linear number of messages in the number of agents, but of exponential size. This is often the main limitation for their practical applicability. Here we present some distributed algorithms for these problems when they are arranged in a tree of agents. The exact algorithm, DCTE, computes the optimal solution but requires messages of size exp(s), where s is a structural parameter. Its approximate version, DMCTE(r), requires smaller messages of size exp(r), r < s, at the cost of computing approximate solutions. It provides a cost interval that bounds the error of the approximation. Using the technique of cost function filtering, we obtain DMCTEf(r). Combining cost function filtering with bound reasoning, we propose DIMCTEf, an algorithm based on repeated executions of DMCTEf(r) with increasing r. DIMCTEf uses messages of previous iterations to decrease the size of messages in the current iteration, which allows to alleviate their high size. We provide evidences of the benefits of our approach on two benchmarks. This work has been partially supported by the project TIN2009-13591-C02-02.

Privacy Loss in Distributed Constraint Reasoning: A Quantitative Framework for Analysis and its Applications

It is critical that agents deployed in real-world settings, such as businesses, offices, universities and research laboratories, protect their individual users' privacy when interacting with other entities. Indeed, privacy is recognized as a key motivating factor in the design of several multiagent algorithms, such as in distributed constraint reasoning (including both algorithms for distributed constraint optimization (DCOP) and distributed constraint satisfaction (DisCSPs)), and researchers have begun to propose metrics for analysis of privacy loss in such multiagent algorithms. Unfortunately, a general quantitative framework to compare these existing metrics for privacy loss or to identify dimensions along which to construct new metrics is currently lacking. This paper presents three key contributions to address this shortcoming. First, the paper presents VPS (Valuations of Possible States), a general quantitative framework to express, analyze and compare existing metrics of privacy loss. Based on a state-space model, VPS is shown to capture various existing measures of privacy created for specific domains of DisCSPs. The utility of VPS is further illustrated through analysis of privacy loss in DCOP algorithms, when such algorithms are used by personal assistant agents to schedule meetings among users. In addition, VPS helps identify dimensions along which to classify and construct new privacy metrics and it also supports their quantitative comparison. Second, the article presents key inference rules that may be used in analysis of privacy loss in DCOP algorithms under different assumptions. Third, detailed experiments based on the VPS-driven analysis lead to the following key results: (i) decentralization by itself does not provide superior protection of privacy in DisCSP/DCOP algorithms when compared with centralization; instead, privacy protection also requires the presence of uncertainty about agents' knowledge of the constraint graph. (ii) one needs to carefully examine the metrics chosen to measure privacy loss; the qualitative properties of privacy loss and hence the conclusions that can be drawn about an algorithm can vary widely based on the metric chosen. This paper should thus serve as a call to arms for further privacy research, particularly within the DisCSP/DCOP arena.