Many routing metrics have been proposed to improve the performance of wireless mesh networks (WMNs). Most of these routing metrics describe intra-flow and inter-flow interference respectively, which brings in adjustable parameters or results in non-isotonicity. However, the adjustable parameters, used to balance the intra-flow and inter-flow interference, are difficult to adapt to the network status, and non-isotonicity makes the design of routing metric complicated. In this paper, we propose an isotonic metric of expected delay (MED) in multi-radio multi-channel (MRMC) WMNs. In particular, MED uses the expected available bandwidth (EAB), which can capture the logical intra-flow and inter-flow interference uniformly, to estimate the delay of the path. In MED, both expected packet transmission delay and expected queuing delay are estimated to capture physical interference, logical interference, load and noise comprehensively. Simulation results show that the proposed metric can improve overall network performance effectively. Index Terms-Wireless mesh networks, routing metric, load, interference, delay I. INTRODUCTION Wireless mesh networks (WMNs) are self-organizing, self-configuring, low-cost networks. WMNs are widely used in different scenarios, such as urban area networks, emergency communication networks, security networks and so on [1]. Since single-radio single-channel WMNs cannot satisfy the increasing demands for the traffic that is growing every day, multi-radio multi-channel (MRMC) WMNs are developed [2]. MRMC WMNs provide each node with multi-radio interfaces, which can improve the capacity of the networks and enhance overall network performance. Due to the limited spectrum resources, interference and congestion have great impact on network performance, so it is important to select a proper path according to the network status. Routing metric is the core of routing protocols, which provides quantifiable values to judge the efficiency of the route [3]. Thus design of routing metrics plays a crucial role for finding efficient routes in the network. In recent years, many researchers focus on studying routing metrics to improve overall network performance of Manuscript