An Efficient Algorithm for Solving Dynamic Complex DCOP Problems

Abstract

Multi Agent Systems and the Distributed Constraint Optimization Problem (DCOP) formalism offer several asynchronous and optimal algorithms for solving naturally distributed optimization problems efficiently. There has been good application of this technology in addressing real world problems in areas like Sensor Networks and Meeting Scheduling. Most of these algorithms however exploit static tree structures and are thus not well suited to modeling and solving problems in rapidly changing domains. Also, while in theory most DCOP algorithms can be extended to handle complex local sub-problems, we argue that this generally results in making their performance sub-optimal, and thus their application less suitable. In this paper we present new measures that emphasize the interconnectedness between each agent's local and interagent sub-problems and use these measures to guide dynamic agent ordering during distributed constraint reasoning. The resulting algorithm, DCDCOP, offers a robust, flexible, and efficient mechanism for modeling and solving dynamic complex problems. Experimental evaluation of the algorithm shows that DCDCOP significantly outperforms ADOPT, the gold standard in search-based DCOP algorithms.

Applicability of Multi-Agent Systems and Constrained Reasoning for Sensor-Based Distributed Scenarios: A Systematic Mapping Study on Dynamic DCOPs

“…In these, studies agents are in control of a set of different elements such as traffic lights on crossroads [ 63 , 65 ] or platforms from railway systems [ 64 ], where dynamics are represented as the variations in the traffic density from circulating vehicles. Other scenarios were Dynamic DCOPs are modelled include collective autonomous vehicles for collision avoidance [ 66 ], human resources reorganisation such as scheduling elective surgery for emergencies in hospitals [ 67 ], distributed controlling of IoT devices and sensors for minimisation of energy consumption in smart homes [ 47 ], and dynamic social simulation through law enforcement problems and market-based mechanisms for dynamic task allocation to agents [ 68 ]. Proposals without scenarios [ 24 , 25 , 39 , 40 , 42 , 43 , 44 , 53 , 54 , 69 , 70 , 71 , 72 , 73 , 74 , 75 ] are the majority of the studies found.…”

“…Other scenarios were Dynamic DCOPs are modelled include collective autonomous vehicles for collision avoidance [ 66 ], human resources reorganisation such as scheduling elective surgery for emergencies in hospitals [ 67 ], distributed controlling of IoT devices and sensors for minimisation of energy consumption in smart homes [ 47 ], and dynamic social simulation through law enforcement problems and market-based mechanisms for dynamic task allocation to agents [ 68 ].…”

Decentralised Control of Intelligent Devices: A Healthcare Facility Study

“…Beyond these first solutions, DCOP algorithms have already been tested in practical scenarios such as travel optimization ( [6], [15]) or planning ( [7]). Among the DCOP algorithms, three are particularly well known: ADOPT [12], DPOP [14], OptAPO [10].…”

“…Both variants are based on a mediator agent, so the resolution is not fully distributed. [7] proposed another algorithm to solve dynamic problems called DCDCOP (Dynamic Complex DCOP) based on a case study of time use optimization in a medical context. This algorithm -mainly based on the addition of a Degree of Unsatisfaction measure -dynamically guides agents through the resolution process.…”

“…The purpose of our system is to warn the medical staff before devices ring, but without increasing the frequency of their interventions. We formulate the problem as a Distributed Constraint Optimization Problem (DCOP) which has been shown to be effective in itinerary optimization [6], [15] and scheduling problems [7]. This decentralized approach has the benefit of distributing the main computations across all available devices.…”