default A survey of multi-agent strategies for ASW applications


A survey of multi-agent strategies for ASW applications. Ferri, Gabriele. CMRE-FR-2015-011. September 2015.

The CMRE multi-static ASW network is composed of fixed nodes and mobile robots working together to achieve common objectives (this kind of network is named in the literature as hybrid). In the ASW littoral scenario the CMRE ASW network offers the promise to provide effective ASW capabilities at a fraction of the cost of traditional assets. To realize an effective network we need to endow every node with autonomy and with the capability to collaborate with one another. Autonomy is crucial for the agents to make effective decisions when communication with other network members and with the command and control centre is limited or impossible. Collaboration between nodes can increase the network robustness and the effectiveness in accomplishing ASW tasks. In this report we start from an analysis of how an effective collaboration between nodes and autonomy can benefit the CMRE ASW network in a littoral surveillance scenario. We then present a thorough review of multi-robot cooperation topics with a special focus on task allocation. Task allocation is the process of allocating tasks to the most suited robots to increase the group's benefit. These studies suggest that a market-based approach can be a viable solution for task allocation. These methods are flexible and provide a good compromise between required communication bandwidth and quality of the task allocation. Guiding principles are presented to design a market-based task allocation scheme tailored to the ASW scenario. In particular, the tasks constituting a complex ASW mission are presented along with a suited utility function. An example of an auction is also presented and discussed. The approach presented can be used as a template to develop auctions also for other tasks of interest.