Multi-player Bandits for Distributed Cognitive Radar

Howard, William W.; Thornton, Charles E.; Martone, Anthony F.; Buehrer, R. Michael

doi:10.1109/radarconf2147009.2021.9455306

Cited by 12 publications

(12 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

Section: B Contributionsmentioning

confidence: 70%

“…This paper extends preliminary versions which were published previously [1], [2]. We previously investigated the available MMAB algorithms, and in [1] studied reward structures which differ between radar nodes. In [2], we analyzed algorithms tailored towards adversarial environments, where the environment has knowledge of the algorithm being used by the radar network, and can pre-select a reward sequence in an attempt to worsen the performance of the network.…”

Section: B Contributionsmentioning

confidence: 70%

“…Varieties of the proposed sequential decision-making problem has been studied in cognitive radio, as well as in cognitive radar [2], [35]. The approach in this work employs a multiplayer multi-arm bandit framework to allow nodes to develop a model of their environment, without relying on a central decision-maker [1]. This means that each node can only access the information it has observed: rewards for each waveform selected, and an indicator for collisions.…”

Section: E Algorithmsmentioning

confidence: 99%

“…In addition, the described CRN uses sensing and agility to avoid causing harmful interference within the network. We discuss the use of a library of orthogonal waveforms, which has been shown to result in both lower ambiguity 1 [6] and reduced mutual interference [7].…”

Section: Introductionmentioning

confidence: 99%

“…The multi-player problem was initially motivated by, among other things, coexistence in cognitive radio applications. We first adapted MMAB algorithms to the cognitive radar application in [1].…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Distributed Online Learning for Coexistence in Cognitive Radar Networks

Howard¹,

Martone²,

Buehrer³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

This work addresses the coexistence problem for radar networks. Specifically, we model a network of cooperative, independent, and non-communicating radar nodes which must share resources within the network as well as with non-cooperative nearby emitters. We approach this problem using online Machine Learning (ML) techniques. Online learning approaches are specifically preferred due to the fact that each radar node has no prior knowledge of the environment nor of the positions of the other radar nodes, and due to the sequential nature of the problem. For this task we specifically select the multi-player multi-armed bandit (MMAB) model, which poses the problem as a sequential game, where each radar node in a network makes independent selections of center frequency and waveform with the same goal of improving tracking performance for the network as a whole. For accurate tracking, each radar node communicates observations to a fusion center on set intervals. The fusion center has knowledge of the radar node placement, but cannot communicate to the individual nodes fast enough for waveform control. Every radar node in the network must learn the behavior of the environment, which includes rewards, interferer behavior, and target behavior. Each independent and identical node must choose one of many waveforms to transmit in each Pulse Repetition Interval (PRI) while avoiding \emph{collisions} with other nodes and interference from the environment. The goal for the network as a whole is to minimize target tracking error, which relies on obtaining high SINR in each time step. Our contributions include a mathematical description of the MMAB framework adapted to the radar network scenario. We conclude with a simulation study of several different network configurations. Experimental results show that iterative, online learning using MMAB outperforms the more traditional sense-and-avoid (SAA) and fixed-allocation approaches.

show abstract

Section: B Contributionsmentioning

confidence: 70%