Co-operative spectrum sensing: Implementation and benchmarking on ANRC Cognitive Radio Testbed

Cognitive Radio is an enabling technology that is expected to address the problem of scarcity of spectrum by intelligent access, mobility and management of spectrum and pave way for efficient usage of the same. Increasingly, it is seen that most research is directed towards building theoretic models for spectrum access and management. Though some of these are backed by simulations, certainty of any method can be proven only by practical implementation on hardware testbeds in real-world environments. Sometimes, assumptions that are made to satisfy simulation requirements are difficult to realize in practical scenarios, thereby, their implementation in real-world is even more challenging. In this paper, taking cue from our earlier implementation work on simple Snapshot ED and CUSUM based algorithms for single node spectrum sensing, we explain the implementation of Sequential change detection algorithm in co-operative mode, the Dual-CUSUM. Single node sensing is not reliable, especially, when the nodes are subjected to shadow fading due to obstacles (hidden node problem). This is mitigated by using co-operative sensing that exploits the inherent spatial diversity and it is seen that with certain fusion rules, Cooperative CUSUM (aka Dual-CUSUM) is better than its counterpart Co-operative Snapshot ED, especially, under low SNR regimes.