Dilian Gurov,Filip Jamroga,Wojciech Jamroga,Mateusz Kamiński,Damian Kurpiewski,Wojciech Penczek,Teofil Sidoruk
Abstract:Model checking of strategic abilities for agents with memory is a notoriously hard problem, and very few attempts have been made to tackle it. In this paper, we present two important steps towards this goal. First, we take the partial-order reduction scheme that was recently proved to preserve individual and coalitional abilities of memoryless agents, and show that it also works for agents with memory. Secondly, we take the Knowledge-Based Subset Construction, that was recently studied for synchronous concurrent games, and adapt it to preserve abilities of memoryful agents in asynchronous MAS.
On the way, we also propose a new execution semantics for strategies in asynchronous MAS, that combines elements of Concurrent Game Structures and Interleaved Interpreted Systems in a natural and intuitive way.
What problem does this paper attempt to address?
### Problems the paper attempts to solve
This paper aims to solve the problem of verifying the strategic capabilities of agents with memory in multi - agent systems (MAS). Specifically, the paper mainly addresses the following two key issues:
1. **Model simplification and knowledge base construction**:
- **Partial Order Reductions (POR)**: The paper extends the partial order reduction scheme, which has been recently proven to be applicable to memory - less agents, to agents with memory, and demonstrates its effectiveness in maintaining individual and group strategic capabilities.
- **Knowledge - Based Subset Construction (KBSC)**: The paper adapts KBSC from synchronous concurrent games to asynchronous MAS to maintain the strategic capabilities of agents with memory.
2. **Improvement of execution semantics**:
- The paper proposes a new strategy execution semantics in asynchronous MAS, which combines elements of concurrent game structures (CGS) and interleaved interpretation systems, providing a natural and intuitive way.
### Explanation of formulas and symbols
To ensure the correctness and readability of formulas, the following are the Markdown - format representations of some key formulas and symbols involved in the paper:
- **Partial Order Reductions**:
\[
(g, in, out, g') \in T \iff \begin{cases}
\text{(*)} & \begin{cases}
out \in Evt \setminus \{\epsilon\}, \\
\forall i \in Agent(out), out \in E_i \wedge (g_i, out, (g')_i) \in T_i, \\
\forall i \in A \setminus Agent(out), (g')_i = g_i
\end{cases} \\
\text{(**)} & \begin{cases}
out = \epsilon, \\
g' = g, \\
\forall \alpha \in Evt \setminus \{\epsilon\}, \exists i \in Agent(\alpha), \alpha \notin E_i
\end{cases}
\end{cases}
\]
- **Knowledge - Based Subset Construction**:
\[
sK = \epsilon\text{-closure}(Q) / \sim_i
\]
where \( \epsilon\text{-closure}(Q) \) represents the \( \epsilon \)-closure of \( Q \), and \( \sim_i \) is the observational equivalence relation with respect to agent \( i \).
### Summary
By introducing new execution semantics and adapting existing partial order reduction and knowledge - based subset construction methods, the paper provides important theoretical and technical support for verifying the strategic capabilities of agents with memory. This not only helps to alleviate the state - space explosion problem, but also provides a feasible method for verifying complex MAS in practical applications.