On Detecting Termination in Cognitive Radio Networks

Abstract: The cognitive radio networks are an emerging wireless communication and computing paradigm. The cognitive radio nodes execute computations on multiple heterogeneous channels in the absence of licensed users (a.k.a. primary users) of those bands. Termination detection is a fundamental and non-trivial problem in distributed systems. In this paper, we propose a termination detection protocol for multi-hop cognitive radio networks where the cognitive radio nodes are allowed to tune to channels that are not currently occupied by primary users and to move to different locations during the protocol execution. The proposed protocol applies credit distribution and aggregation approach and maintains a new kind of logical structure, called the virtual tree-like structure. The virtual tree-like structure helps in decreasing the latency involved in announcing termination. Unlike conventional tree structures, the virtual tree-like structure does not require a specific node to act as the root node that has to stay involved in the computation until termination announcement; hence, the root node may become idle soon after finishing its computation. Also, the protocol is able to detect the presence of licensed users and announce strong or weak termination, whichever is possible.