MAC design for WiFi infrastructure networks: a game-theoretic approach

Ilenia Tinnirello, Laura Giarré, Giovanni Neglia

Abstract

In WiFi networks, mobile nodes compete for accessing a shared channel by means of a random access protocol called Distributed Coordination Function (DCF). Although this protocol is in principle fair, since all the stations have the same probability to transmit on the channel, it has been shown that unfair behaviors may emerge in actual networking scenarios because of non-standard configurations of the nodes. Due to the proliferation of open source drivers and programmable cards, enabling an easy customization of the channel access policies, we propose a game-theoretic analysis of random access schemes. Assuming that each node is rational and implements a best response strategy, we show that efficient equilibria conditions can be reached when stations are interested in both uploading and downloading traffic. More interesting, these equilibria are reached when all the stations play the same strategy, thus guaranteeing a fair resource sharing. When stations are interested in upload traffic only, we also propose a mechanism design, based on an artificial dropping of layer-2 acknowledgments, to force desired equilibria. Finally, we propose and evaluate some simple DCF extensions for practically implementing our theoretical findings.

Introduction

The Distributed Coordination Function (DCF) is the cornerstone of WiFi Medium Access Control (MAC), designed to provide fair channel access through a carrier-sense multiple access with collision avoidance (CSMA/CA) mechanism. However, the protocol's reliance on voluntary compliance makes it vulnerable. As open-source drivers and programmable network interface cards become ubiquitous, it is increasingly easy for "selfish" users to modify their MAC parameters to gain an unfair share of bandwidth. This paper applies game theory to redesign WiFi access, ensuring that fairness becomes the most profitable strategy for all participants.

The Problem: Cheaters and Bottlenecks

Standard WiFi networks suffer from two primary structural weaknesses:

  1. The Cheater’s Advantage: DCF uses a random backoff timer called the Contention Window (CW). A compliant node waits for a random duration before transmitting to avoid collisions. A selfish node can simply hard-code a smaller CW, effectively "cutting in line" and starving compliant nodes of airtime.
  2. The Downlink Bottleneck: In infrastructure mode, an Access Point (AP) handles all traffic destined for multiple stations. Because the AP must compete for the channel using the same rules as a single mobile station, the aggregate download bandwidth for the entire network is often reduced to the upload capacity of a single user. As the number of stations increases, download speeds plummet while upload speeds remain disproportionately high.

Methodology: Game Theory and Mechanism Design

The authors model the WiFi channel as a non-cooperative game where the players are the mobile stations and the AP. Each player's strategy is defined by their transmission probability ($\tau$).

Utility Functions

The core of the model is the utility function, which represents a node's goal. The authors assume a rational node values both upload and download throughput. For example, a web browser cannot function if it can send requests (upload) but cannot receive the resulting data (download). This interdependence is the key to incentivizing cooperation.

Mechanism Design

To manage the network, the authors propose two interventions where the AP acts as an arbitrator:

  • AP Tuning: The AP dynamically calculates and adopts an optimal transmission probability to guide the network toward a fair Nash Equilibrium.
  • ACK Suppression: In scenarios where nodes only care about uploading (unidirectional traffic), the AP can "police" the network. If a node transmits too aggressively, the AP intentionally drops the Layer-2 Acknowledgement (ACK). The node, perceiving a collision, is forced into an exponential backoff, effectively slowing it down without requiring any software changes on the client side.

Performance Evaluation

The proposed models were evaluated using a custom C++ simulation platform across various physical layer standards, including 802.11b and 802.11g/n. The simulations tested networks with 2 to 20 stations under dynamic conditions where nodes joined or left the network.

Key Findings

  • Throughput Stability: In a 20-station environment, the proposed scheme maintained a total network throughput of approximately 5 Mbps. In contrast, standard DCF throughput dropped to 3.8 Mbps due to collision overhead and downlink starvation.
  • Fairness and Neutralization: The system successfully equalized upload and download traffic when configured to do so. Furthermore, even when a "cheater" used an extremely aggressive contention window (e.g., CW=8), the AP’s ACK suppression successfully forced the cheater's throughput down to the level of compliant nodes.
  • Self-Regulating Incentives: The research demonstrates that if a node requires download traffic, its "best response" is not to flood the uplink. Instead, it must leave enough airtime for the AP to transmit, creating a natural incentive for fair play.

Conclusion and Future Work

By treating WiFi access as a game-theoretic problem, this paper provides a robust framework for enforcing fairness in infrastructure networks. The use of ACK suppression is particularly powerful as it exploits the existing logic of the WiFi protocol to regulate selfish behavior.

However, implementation challenges remain. ACK suppression requires microsecond-level precision, likely necessitating FPGA-level access to the AP's hardware. Additionally, while the model assumes "saturated" traffic conditions (where nodes always have data to send), future research is needed to adapt these strategies for the bursty, non-saturated traffic patterns typical of real-world usage.

Summary updated

Abstract

In WiFi networks, mobile nodes compete for accessing a shared channel by means of a random access protocol called Distributed Coordination Function (DCF). Although this protocol is in principle fair, since all the stations have the same probability to transmit on the channel, it has been shown that unfair behaviors may emerge in actual networking scenarios because of non-standard configurations of the nodes. Due to the proliferation of open source drivers and programmable cards, enabling an easy customization of the channel access policies, we propose a game-theoretic analysis of random access schemes. Assuming that each node is rational and implements a best response strategy, we show that efficient equilibria conditions can be reached when stations are interested in both uploading and downloading traffic. More interesting, these equilibria are reached when all the stations play the same strategy, thus guaranteeing a fair resource sharing. When stations are interested in upload traffic only, we also propose a mechanism design, based on an artificial dropping of layer-2 acknowledgments, to force desired equilibria. Finally, we propose and evaluate some simple DCF extensions for practically implementing our theoretical findings.
BibTeX
@article{tinnirello2011mac,
  author = {Ilenia Tinnirello and Laura Giarré and Giovanni Neglia},
  title = {{MAC design for WiFi infrastructure networks: a game-theoretic approach}},
  journal = {IEEE Trans. Wireless Communications},
  year = {2011},
  url = {https://openreview.net/forum?id=lmYRrvUwCl5},
}
Figure