A Fault Tolerant Scheme for Ad-hoc Networks

Different from the traditional distributed computing system, ad-hoc networks have many new features. The transient failure probability of the computing process increases with the increase of system scale in ad-hoc networks. To reduce the loss of computation upon the transient process failure event, a fault tolerant scheme for Ad-hoc networks is proposed in this paper. The proposal can seamlessly cooperate with the cluster-based multi-channel ad-hoc network system. A low overhead of the communication between the cluster head and related ordinary members is required. The recovery process has no domino effect and the failure process can rollback from its latest local consistent fault tolerant information. By the discussion and contrast, the proposed scheme is able to recover quickly upon the transient failure while only a low additional overhead is required. Introduction Due to the features including dynamic mobility, celerity of setting, autonomy, flexibility of topology and equivalence, the mobile computing network has broad application foreground [1]. An ad-hoc network is a collection of wireless mobile nodes forming a temporary network without the aid of any extra static power processing centralized administration acting as the mobile supporting stations [2-3]. This type of network is not supported by a wired infrastructure like a conventional cellular system [4]. Compared to the traditional wired distributed computing system, the topology of such networks is very dynamic because of host mobility, packet loss, interference, and low power. There are several other features of ad-hoc networks different from wired networks, such as low bandwidth, limited storage space and low computing capacity[5]. Cluster-based multi-channel ad-hoc network. As shown in Figure 1, the system consists of a set of mobile hosts (MH), which communicate with each other through wireless channels [6-7]. We assume that wireless channels are all FIFO order.