We present a simulation-based approach for solution of mean field games (MFGs), using the framework of empirical game-theoretical analysis (EGTA). Our primary method employs a version of the double oracle, iteratively adding strategies based on best response to the equilibrium of the empirical MFG among strategies considered so far. We present Fictitious Play (FP) and Replicator Dynamics as two subroutines for computing the empirical game equilibrium. Each subroutine is implemented with a query-based method rather than maintaining an explicit payoff matrix as in typical EGTA methods due to a representation issue we highlight for MFGs. By introducing game model learning and regularization, we significantly improve the sample efficiency of the primary method without sacrificing the overall learning performance. Theoretically, we prove that a Nash equilibrium (NE) exists in the empirical MFG and show the convergence of iterative EGTA to NE of the full MFG with either subroutine. We test the performance of iterative EGTA in various games and show that it outperforms directly applying FP to MFGs in terms of iterations of strategy introduction.

Majority of aircraft under the Urban Air Mobility (UAM) concept are expected to be of the electric vertical takeoff and landing (eVTOL) vehicle type, which will operate out of vertiports. While this is akin to the relationship between general aviation aircraft and airports, the conceived location of vertiports within dense urban environments presents unique challenges in managing the air traffic served by a vertiport. This challenge becomes pronounced within increasing frequency of scheduled landings and take-offs. This paper assumes a centralized air traffic controller (ATC) to explore the performance of a new AI driven ATC approach to manage the eVTOLs served by the vertiport. Minimum separation-driven safety and delays are the two important considerations in this case. The ATC problem is modeled as a task allocation problem, and uncertainties due to communication disruptions (e.g., poor link quality) and inclement weather (e.g., high gust effects) are added as a small probability of action failures. To learn the vertiport ATC policy, a novel graph-based reinforcement learning (RL) solution called "Urban Air Mobility- Vertiport Schedule Management (UAM-VSM)" is developed. This approach uses graph convolutional networks (GCNs) to abstract the vertiport space and eVTOL space as graphs, and aggregate information for a centralized ATC agent to help generalize the environment. Unreal Engine combined with Airsim is used as the simulation environment over which training and testing occurs. Uncertainties are considered only during testing, due to the high cost of Mc sampling over such realistic simulations. The proposed graph RL method demonstrates significantly better performance on the test scenarios when compared against a feasible random decision-making baseline and a first come first serve (FCFS) baseline, including the ability to generalize to unseen scenarios and with uncertainties.

Over these years, multi-agent reinforcement learning has achieved remarkable performance in multi-agent planning and scheduling tasks. It typically follows the self-play setting, where agents are trained by playing with a fixed group of agents. However, in the face of zero-shot coordination, where an agent must coordinate with unseen partners, self-play agents may fail. Several methods have been proposed to handle this problem, but they either take a lot of time or lack generalizability. In this paper, we firstly reveal an important phenomenon: the zero-shot coordination performance is strongly linearly correlated with the similarity between an agent's training partner and testing partner. Inspired by it, we put forward a Similarity-Based Robust Training (SBRT) scheme that improves agents' zero-shot coordination performance by disturbing their partners' actions during training according to a pre-defined policy similarity value. To validate its effectiveness, we apply our scheme to three multi-agent reinforcement learning frameworks and achieve better performance compared with previous methods.

Visual imitation learning enables reinforcement learning agents to learn to behave from expert visual demonstrations such as videos or image sequences, without explicit, well-defined rewards. Previous research either adopted supervised learning techniques or induce simple and coarse scalar rewards from pixels, neglecting the dense information contained in the image demonstrations. In this work, we propose to measure the expertise of various local regions of image samples, or called \textit{patches}, and recover multi-dimensional \textit{patch rewards} accordingly. Patch reward is a more precise rewarding characterization that serves as a fine-grained expertise measurement and visual explainability tool. Specifically, we present Adversarial Imitation Learning with Patch Rewards (PatchAIL), which employs a patch-based discriminator to measure the expertise of different local parts from given images and provide patch rewards. The patch-based knowledge is also used to regularize the aggregated reward and stabilize the training. We evaluate our method on DeepMind Control Suite and Atari tasks. The experiment results have demonstrated that PatchAIL outperforms baseline methods and provides valuable interpretations for visual demonstrations.

This work extends an existing virtual multi-agent platform called RoboSumo to create TripleSumo -- a platform for investigating multi-agent cooperative behaviors in continuous action spaces, with physical contact in an adversarial environment. In this paper we investigate a scenario in which two agents, namely'Bug' and'Ant', must team up and push another agent'Spider' out of the arena. To tackle this goal, the newly added agent'Bug' is trained during an ongoing match between'Ant' and'Spider'. 'Bug' must develop awareness of the other agents' actions, infer the strategy of both sides, and eventually learn an action policy to cooperate. The reinforcement learning algorithm Deep Deterministic Policy Gradient (DDPG) is implemented with a hybrid reward structure combining dense and sparse rewards. The cooperative behavior is quantitatively evaluated by the mean probability of winning the match and mean number of steps needed to win.

We initiate the study of strategic behavior in screening processes with multiple classifiers. We focus on two contrasting settings: a conjunctive setting in which an individual must satisfy all classifiers simultaneously, and a sequential setting in which an individual to succeed must satisfy classifiers one at a time. In other words, we introduce the combination of strategic classification with screening processes. We show that sequential screening pipelines exhibit new and surprising behavior where individuals can exploit the sequential ordering of the tests to zig-zag between classifiers without having to simultaneously satisfy all of them. We demonstrate an individual can obtain a positive outcome using a limited manipulation budget even when far from the intersection of the positive regions of every classifier. Finally, we consider a learner whose goal is to design a sequential screening process that is robust to such manipulations, and provide a construction for the learner that optimizes a natural objective.

Communication in multi-agent reinforcement learning has been drawing attention recently for its significant role in cooperation. However, multi-agent systems may suffer from limitations on communication resources and thus need efficient communication techniques in real-world scenarios. According to the Shannon-Hartley theorem, messages to be transmitted reliably in worse channels require lower entropy. Therefore, we aim to reduce message entropy in multi-agent communication. A fundamental challenge is that the gradients of entropy are either 0 or infinity, disabling gradient-based methods. To handle it, we propose a pseudo gradient descent scheme, which reduces entropy by adjusting the distributions of messages wisely. We conduct experiments on two base communication frameworks with six environment settings and find that our scheme can reduce message entropy by up to 90% with nearly no loss of cooperation performance.

This paper puts forth a new, reconfigurable intelligent surface (RIS)-assisted, uplink, user-centric cell-free (UCCF) system managed with the assistance of a digital twin (DT). Specifically, we propose a novel learning framework that maximizes the sum-rate by jointly optimizing the access point and user association (AUA), power control, and RIS beamforming. This problem is challenging and has never been addressed due to its prohibitively large and complex solution space. Our framework decouples the AUA from the power control and RIS beamforming (PCRB) based on the different natures of their variables, hence reducing the solution space. A new position-adaptive binary particle swarm optimization (PABPSO) method is designed for the AUA. Two twin-delayed deep deterministic policy gradient (TD3) models with new and refined state pre-processing layers are developed for the PCRB. Another important aspect is that a DT is leveraged to train the learning framework with its replay of channel estimates stored. The AUA, power control, and RIS beamforming are only tested in the physical environment at the end of selected epochs. Simulations show that using RISs contributes to considerable increases in the sum-rate of UCCF systems, and the DT dramatically reduces overhead with marginal performance loss. The proposed framework is superior to its alternatives in terms of sum-rate and convergence stability. Y. Cui and T. Lv are with the School of Information and Communication Engineering, Beijing University of Posts and Telecommunications (BUPT), Beijing 100876, China (e-mail: {cuiyingping, lvtiejun,}@bupt.edu.cn).

Value function factorization methods have become a dominant approach for cooperative multiagent reinforcement learning under a centralized training and decentralized execution paradigm. By factorizing the optimal joint action-value function using a monotonic mixing function of agents' utilities, these algorithms ensure the consistency between joint and local action selections for decentralized decision-making. Nevertheless, the use of monotonic mixing functions also induces representational limitations. Finding the optimal projection of an unrestricted mixing function onto monotonic function classes is still an open problem. To this end, we propose ReMIX, formulating this optimal projection problem for value function factorization as a regret minimization over the projection weights of different state-action values. Such an optimization problem can be relaxed and solved using the Lagrangian multiplier method to obtain the close-form optimal projection weights. By minimizing the resulting policy regret, we can narrow the gap between the optimal and the restricted monotonic mixing functions, thus obtaining an improved monotonic value function factorization. Our experimental results on Predator-Prey and StarCraft Multiagent Challenge environments demonstrate the effectiveness of our method, indicating the better capabilities of handling environments with non-monotonic value functions.

The determination of the computational complexity of multi-agent pathfinding on directed graphs (diMAPF) has been an open research problem for many years. While diMAPF has been shown to be polynomial for some special cases, only recently, it has been established that the problem is NP-hard in general. Further, it has been proved that diMAPF will be in NP if the short solution hypothesis for strongly connected directed graphs is correct. In this paper, it is shown that this hypothesis is indeed true, even when one allows for synchronous rotations.