On the Detection Time of a Primary Network Using Fusion Rules in a Cognitive WLAN Network

In this paper, we are interested in a cognitive secondary wireless local area network (WLAN) which operates in the same or overlapping spectrum and coverage with a primary wireless network. The protection of primary network is quantified by the detection time, which defines the number of superframes required for the secondary network to recognize the re-appearance of the primary network. We obtain the mean and distribution of the detection time of the primary network using the cognitive WLAN, which has an access point (AP) and a number of stations, either sleeping or active. All of these stations are within the detection coverage of the base station (BS) of the primary network. If the BS of the primary network becomes active, the WLAN must be able to detect this activity and vacates the channel within certain amount of time. The mean probability of detection of the BS by each station is modeled as a function of the distance between the BS and the station within the coverage of the AP. Three fusion rules, including OR-fusion rule, AND-fusion rule and Majority rule, are used by the WLAN AP to fuse the sensing results obtained by the active WLAN stations. The mean and the distribution of the detection time of the primary network are derived by generically modeling the detection time by an absorbing discrete time Markov chain. The tradeoffs among different fusions rules and the distance between the BS and AP are investigated.