Demonstration of Sequence Detection Algorithms for Dynamic Spectrum Access Networks

Spectrum sensing is a critical requirement for proposed Dynamic Spectrum Access networks both to avoid causing interference to primary users and to maximize the throughput of the secondary network. Traditional sensing algorithms are performance limited by the length of time they can observe the primary user´s transmission. We implement for demonstrative purposes a number of sequence detection spectrum sensing algorithms that take into account the transitions in the primary user´s channel access, resulting in better sensing performance. The demonstration consists of primary and secondary data links that are both streaming video in the same frequency band, forcing the secondary link to opportunistically access the spectrum. Various sensing algorithms are selectable and configurable at the secondary transmitter, including multiple sequence detection algorithms and energy detection. The superior performance of the sequence detection algorithms is evident at the receivers through a number of metrics, including video quality, plots of the historical data rate, and estimates of the detection and false alarm probabilities.