Empirical Assessment of CSS Techniques: Efficiency, Reliability, and Throughput Trade-offs

As the number of wireless communication services increases exponentially, there is an acute shortage of available frequency spectrum. Allocation of spectrum statically and conventionally leads to the inefficient use of spectrum because some licensed frequency bands become underutilized, while the other unlicensed bands may be congested. A solution proposed in such cases is Cognitive Radio (CR), which offers dynamic spectrum allocation based on intelligent spectrum sensing. There are different techniques of sensing, and one such technique is Cooperative Spectrum Sensing (CSS) in which the information about sensing is shared by multiple CR terminals. In this paper, we analyze the performance of various cooperative spectrum sensing schemes in a cognitive radio environment, viz., centralized, distributed, relay aided and external cooperative sensing. We conduct performance comparison in terms of important factors, namely probability of detection (Pd), probability of false alarm (Pf) and Signal-to- Noise Ratio (SNR). Also, we explore the bandwidth-efficient techniques of cooperative sensing using censoring and quantization techniques. The simulations clearly show that cooperative sensing outperforms non-cooperative sensing by a fair margin but at the cost of little bandwidth efficiency.