This paper proposes two novel hidden node problem aware routing metrics for wireless local area network (WLAN) mesh networks. To select the path that is least affected by the serious hidden node problem, we propose two routing metrics, MNn and MPCP, that take into account the number of neighboring nodes (Nn) and the packet collision probability (PCP), respectively. The PCP is estimated from neighbor information that is periodically gathered as state announcement packets, which include the transmission time ratio and the neighbor list. Simulation results show that the first proposed MNn routing metric tends to be less effective as the number of WLAN nodes increases, i.e., the mesh network becomes denser. On the other hand, with an acceptable increased in the control overhead in the mesh network due to the neighbor information, the second proposed MPCP routing metric improves the number of allowable concurrent voice over Internet Protocol (VoIP) calls and the user datagram protocol (UDP) data throughput compared to the MNn metric. The MPCP also provides better performance than the other conventional routing metrics, the hop count, and the Airtime proposed in IEEE 802.11s.

The dynamic channel selection mechanism used in existing multi-channel MAC protocols selects an idle data channel based on channel usage information from one-hop neighbor nodes. However, this method can cause multi-channel hidden node problem in multi-hop wireless networks. This letter proposes a new approach to channel selection. Nodes snoop data channels during idle times and then select an idle data channel within the carrier sensing range using both the snooping results and the channel usage information. Our simulation results verify that the proposed channel selection approach can effectively avoid the multi-channel hidden node problem and improve the networkwide performance.