Dynamic spectrum access with packet size adaptation and residual energy balancing for energy-constrained cognitive radio sensor networks

We demonstrate an improvement in energy efficiency and network lifetime for the cluster-based multi-channel cognitive radio sensor network (CRSN). The improvement roots from two techniques proposed in dynamic spectrum access. The first technique exploits packet size adaptation: varying the packet size to adapt the transmission over the state-varying channel. This is to efficiently utilize the battery of sensors by having the most appropriately sized packets successfully transmitted, in accordance with the instantaneous channel conditions. The second technique focuses on channel assignment with awareness of the residual energy of sensors, such that sensors can spend their energy in a balanced way. This helps to prolong the network lifetime, compared to the random channel pairing approach. As all those techniques rely on the estimates of channel states and their performance is tied with the estimation accuracy, we theoretically derive a polynomial-time resolvable expression for the maximum-likelihood (ML) estimator PMF function. In light of this expression, the impact of channel estimation accuracy on network performance is thereby illustrated.