Experimental Comparison of Decentralized Task Allocation Algorithms Under Imperfect Communication

Nayak, Sharan; Otte, Michael W.; Yeotikar, Suyash; Carrillo, Estefany; Rudnick-Cohen, Eliot; Jaffar, Mohamed Khalid M; Patel, Ruchir; Azarm, Shapour; Herrmann, Jeffrey W.; Xu, Huan

doi:10.1109/lra.2019.2963646

Cited by 36 publications

(27 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The communication estimate c ei is then mapped to a discrete communication level. Nayak et al [2] showed that the performance ranking of the task allocation algorithms tested remains the same at high communication levels for the visit and search & visit scenarios. It is only when communication drops substantially that this ranking changes and remains constant for lower levels of communication.…”

Section: A Environmentmentioning

confidence: 99%

“…It is only when communication drops substantially that this ranking changes and remains constant for lower levels of communication. For this reason, we define only two discrete communication levels based on the analysis presented in [2]: high (H) for c t < c ei ≤ 1 and low (L) for 0 ≤ c ei ≤ c t . The threshold value c t is specified experimentally by mapping the sensitivity threshold values tested in [2] to either high or low communication based on the sensitivity threshold value at which the change in ranking of the algorithms was observed.…”

Section: A Environmentmentioning

confidence: 99%

“…However, communication is unreliable in realistic environments and outside the control of the multi-robot team, making the coordination of tasks more challenging. Otte et al [1] and Nayak et al [2] have shown that different distributed task allocation algorithms perform better, relative to each other, at different communication quality levels.…”

Section: Introductionmentioning

confidence: 99%

“…3) We validate the proposed approach via a large number of experiments across multiple types of scenarios. In this paper, the multi-agent metareasoning policy is built from a set of decentralized task allocation methods for which performance profiles under varying levels of communication were obtained in [2]. This set includes: the Consensus Based Auction Algorithm (CBAA) [14], the Asynchronous Consensus Based Bundle Algorithm (ACBBA) [15] [16], the Decentralized Hungarian Based Algorithm (DHBA) [17], the Hybrid Information and Plan Consensus (HIPC) algorithm [18] [19] and the Performance Impact (PI) algorithm [20].…”

Section: Introductionmentioning

confidence: 99%

“…Changes in signal attenuation (i.e., communication quality) over time are modeled by varying the path loss exponent parameter. We adopt the Rayleigh fading model to simulate imperfect communication environments as it comes closer to modeling realistic environments compared to other models such as the Bernoulli or Gilbert-Elliot model models [2].…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Communication-Aware Multi-Agent Metareasoning for Decentralized Task Allocation

et al. 2021

Self Cite

View full text Add to dashboard Cite

Metareasoning refers to reasoning about one's own decision making process. This paper considers metareasoning about the decision making process in multi-agent settings. We present a multiagent metareasoning approach that enables a multi-agent team to select which task allocation algorithm to use as a function of changing communication quality level. Given a set of multi-agent task allocation algorithms, we synthesize a policy that prescribes the best algorithm to use among a predefined set of algorithms for a given communication level. Since each agent in the team runs the same policy, the team (or a part of the team) will collectively switch between task allocation algorithms as a function of the observed level of communication. We apply reactive synthesis to generate the policy from high-level specifications written in Linear Temporal Logic encoding the agents' switching behavior with respect to the state of the environment. We perform experiments in simulation to identify the best performing algorithms under different communication levels. The communication environment is modeled using the Rayleigh fading model and communication estimation is done through the exchange of heartbeat messages among agents. We test our metareasoning policy in three types of scenarios: search & rescue, fire monitoring, and ship protection scenarios. For each scenario, we demonstrate that our policy achieved better performance with respect to either max distance traveled, max number of transmitted messages or both compared to running any single algorithm.

show abstract

Section: A Environmentmentioning

confidence: 99%

Section: A Environmentmentioning

confidence: 99%