Research On Wireless Access Network Channel Access Technology Based On Packet Detection

In this dissertation, we focus on wireless sensor networks for group-based event detection where the observed area is covered by concurrently contending N sensor nodes. Upon an event occurrence, each sensor node generates a detection report and tries to transmit it to a central controller following some specific channel access schemes. Such an event detection is employed to improve the report accuracy and avoid the false alarm, and is said to be successful if the controller receives m reports from different sensors (m≤N) within a predefined time interval since the event occurrence. We study the performance of group-based event detection for different types of channel access schemes, including IEEE 802.15.4 slotted CSMA/CA, slotted ALOHA and protocol sequences. Our analytical results are validated against simulation results obtained through NS-2 and Matlab.First, given different contention window settings, we derive the successful probability of group-based event detection which employs IEEE 802.15.4 CSMA/CA as its access scheme. The analysis is following a non-stationary Markov chain model.Second, the performance of group-based event detection is investigated for the ALOHA scheme. Moreover, we present how to seek the optimal accessing probability for maximizing the successful probability of group-based event detection.Third, we are interested in the group-based event detection with the CRT sequences based scheme. The cumulative distribution function of its receive delay is derived.