Throughput performance of tree collision resolution algorithms with capture
The effect of capture on tree algorithms in a slotted ALOHA broadcasting network is investigated. Two receiver models are considered: Feedback With Capture (FWC) and Feedback Without Capture (FWOC). In FWC, the receiver is able to distinguish between a success and capture slot, whereas in FWOC it...
Main Author: | |
---|---|
Language: | English |
Published: |
2009
|
Online Access: | http://hdl.handle.net/2429/5606 |
Summary: | The effect of capture on tree algorithms in a slotted ALOHA broadcasting network is
investigated. Two receiver models are considered: Feedback With Capture (FWC) and
Feedback Without Capture (FWOC). In FWC, the receiver is able to distinguish between
a success and capture slot, whereas in FWOC it is not. For each model, two collision
resolution schemes are examined. The throughput performances of the four schemes
using the discrete and continuous capture models are obtained and compared.
In the discrete capture model, the transmitters are divided into groups. Only packets
from transmitters in a more dominant group have a chance to be captured. The continuous
capture model is used to examine the throughput performance on the inbound
(mobile-to-base station) channel in a packet radio system consisting of a central base station
and a number of mobile user terminals. The use of a dynamic tree (DT) algorithm
in a continuous capture environment is also investigated.
Expressions for finding the average length of a collision resolution interval (CRI) for
each scheme using the static binary tree are derived. These average lengths are used
to determine the throughput. Numerical procedures are described which use the static
binary tree results in the dynamic tree to estimate the throughput. Simulations are used
to verify the maximum achievable throughput in DT which also give plots of the average
packet delay against the arrival rate. |
---|