Abstract:In today's dynamic and interconnected world, resource constraints pose significant challenges across various domains, ranging from networks, logistics and manufacturing to project management and optimization, etc. Resource-constrained problems (RCPs) represent a class of complex computational problems that require efficient allocation and utilization of limited resources to achieve optimal outcomes. This thesis aims to delve into such problems involving multiple agents, where agents aim to enhance their own payoffs, or a neutral moderator aims to maximise the system revenue while distributing the resources appropriately among all agents. In the former type of problems, agents may seek collaboration to achieve higher individual shares, resulting in a cooperative game with competition, i.e., co-opetition. Cooperative and non-cooperative game theory tools are utilized to analyze such games. On the other hand, for the latter kind of problems, we use tools from optimization and Markov decision processes.
What problem does this paper attempt to address?
### Problems Addressed by the Paper
This paper aims to address the issue of resource allocation among multiple self-interested agents in resource-constrained environments, particularly in scenarios where cooperation and competition coexist (i.e., "coopetition"). Specifically, the paper explores the following major problems:
1. **Coalition Formation Games in Online Auctions**:
- In an online auction environment, agents bid using proportional allocation algorithms (such as the Kelly mechanism) to obtain a larger share of the spectrum. Additionally, procurement costs proportional to the bids are considered. Unlike previous studies, this paper explores the possibility of cooperation among agents, leading to a partition form coalition formation game. The study finds that when there are more than 4 symmetric agents with the same reputation factor, the partition consisting of all individual agents is the only stable partition that can resist unilateral deviations; however, no partition can resist coalition deviations. Asymmetry among agents (based on reputation factors) increases the number of stable partitions that can resist unilateral deviations; when the asymmetry reaches a certain threshold, all partitions become stable; interestingly, only large coalitions can resist coalition deviations. The paper also explores the cost of non-cooperation, measured using the Price of Anarchy.
2. **Coalition Formation Games in Loss Queuing Systems**:
- The study examines a queuing system with multiple service providers, each having a fixed number of servers. Unlike previous research assumptions, this paper assumes a fixed shared customer pool, with customers allocated based on the number of servers each service provider has, using the well-known Wardrop equilibrium. This leads to competition among service providers for market share. The possibility of cooperation among service providers is again considered. The results show that no partition can resist coalition deviations (i.e., the core is empty), leading to the introduction of new stability concepts. Using these new concepts, it is found that the most prominent stable partition is an oligopoly.
3. **Opportunistic Fair Scheduler in Future Networks Based on User Location**:
- A system consisting of a base station and multiple mobile/static users is considered. The base station uses millimeter waves (mmWaves) for data transmission, requiring beam alignment with the terminal users' directions. Unlike existing literature that selects only one user for beam alignment and data transmission, this paper designs a dual scheduler to ensure: (i) dynamic updating of user location information, and (ii) optimal channel allocation to various users at different time slots. The goal is to maintain accurate position estimates for each user, enabling multi-user beam alignment. This scheduler optimizes the known alpha-fairness objective function, considering the quality of user location information. The proposed scheduler has a near-closed-form expression, selecting the best option from a finite set. For the case of 2 users, each selection requires solving an N-1 dimensional deterministic equation. These schedulers significantly outperform previously proposed mmWave schemes and offer flexibility to achieve any desired level of fairness.
### Major Contributions
- **Cooperative Game Theory**: Extends solution concepts and stability concepts to partition form games and introduces new stability concepts.
- **Queueing Theory**: Considers the case where service providers segment the market through quality of service, showing results different from existing literature.
- **Markov Decision Process**: Studies a new average cost Markov decision process, optimizing a finite number of average utility functions.
- **Opportunistic Scheduler**: Extends known opportunistic schedulers to future networks, making dual decisions.