Multi-hop coverage extension can be utilized as a feasible approach to facilitating uncovered users to get Internet service in public area WLANs. In this paper we introduce a relay-union network (RUN), which refers to a public area WLAN in which users often wander in the same area and have the ability to provide data forwarding services for others. We develop a RUN framework to model the cost of providing forwarding services and the utility obtained by gaining services. The objective of the RUN is to maximize the total Quality of Cooperation (QoC) of users in the RUN. Two optimal bandwidth allocation schemes are proposed for both free and dynamic bandwidth demand models. To make our scheme more pragmatic, we then consider a more practical scenario in which the bandwidth capacity of the relays and the minimum demand of the clients are bounded. We prove that the problems under both the single relay and the multi-relay scenario are NP-hard. Three heuristic algorithms are proposed to deal with bandwidth allocation and relay-client association. We also propose a distributed signaling protocol and divide the centralized MRMC algorithm into three distributed ones to better adapt for real network environment. Finally, extensive simulations demonstrate that our RUN framework can significantly improve the efficiency of cooperation in the long term.

Public area WLANs stand for WLANs deployed in public areas such as classroom and office buildings to provide Internet connections. Nevertheless, such a service may not be always available because of limited AP coverage, poor signal strength, or password authentication. Multi-hop access is a feasible approach to facilitate users without direct AP accesses to resort to other online users as relays for data forwarding. This paper employs credit-exchange for multi-hop access in public area WLANs to encourage users to cooperate, and proposes a complete pricing framework. We first investigate a revenue model to define the profit of a relay. Next we point out that cutoff bandwidth allocation is a crucial issue in pricing strategy. Optimal bandwidth allocation schemes are then proposed for two bandwidth demand models. Following that we consider a more practical scenario where the relay's bandwidth capacity and the client's bandwidth demand are bounded, and propose two heuristic algorithms SRMC and MRMC to compute bandwidth allocation and/or relay-client association. Extensive simulation study has been performed to validate our design.