Analysis of Search Schemes in Cognitive Radio

Abstract

Abstract-Cognitive Radio systems face an important challenge -fast and reliable channel searching to enable secondary users to optimize available spectral resources. We revisit conventional urn models for channel occupancy and analyze the performance of several search schemes in terms of the mean time to detection. In particular, we focus on correlated Markov models for bin occupancy and highlight the performance of n-step serial search (nSS) algorithm Index Terms-Cognitive Radio, channel model, search scheme

A survey on MAC protocols for cognitive radio networks

“…The performance of several such spectrum search schemes are investigated in [17]. An interesting approach is the consideration of correlated spectrum band occupancy models, in which it is more likely to detect a PU transmission in the neighboring spectrum of a band that is already known to be occupied.…”

“…In addition, PUs may use several spectrum bands at a time, depending upon the way these bands are structured for use by the licensed users. Random and serial search schemes are investigated in [17], and a general n-step serial search scheme is proposed at the best solution.…”

Modeling dynamics of complex wireless networks

“…To date, only a few efforts [2], [25]- [26] have been reported that capture the requisite temporal and frequency-dependent dynamics of primary user occupancy. Most work to date within the academic cognitive radio community [3], [27] assume hypothetical channel idleness models, notably as a set of Bernoulli, independent, and identically distributed (i.i.d.) variables for channel idleness probabilities (CIPs).…”

“…In this case, by increasing the number of allocated access points (m out of M), the delay is expected to be lower resulting in higher value of the first term in (27). On the other hand, if each new access point is charging the operator for this service, the operator will be interested to keep the cost χ lower which results into lower p m .…”

“…8, we present the utility defined by (27) versus p m for τ 0 = 4.22 and different values of χ. Parameter m is the number of terminals turned to be access points. We can see that there is a clear maximum in the utility for the optimum value of p m , obtained as a compromise between m and the cost χ.…”