Abstract:
There is a strong likelihood that future wireless communication systems will employ packetised
transmission techniques. Such an approach would provide a number of benefits, including the
ability to integrate with emerging wired systems based on Asynchronous Transfer Mode (ATM).
In addition, packet-based wireless networks would allow for diverse traffic types and in
particular, speech and data traffic, to be integrated more readily.
This thesis presents theoretical models and performance studies for various interference limited,
packet-based, speech and data wireless communication systems. Two packetised random access
protocols are considered for investigation, namely Slotted ALOHA (S-ALOHA) and Packet
Reservation Multiple Access (PRMA). The S-ALOHA protocol is designed specifically for data
traffic, whereas the PRMA protocol is optimised for speech traffic, with the ability to also cater
for data traffic.
Previously published research into S-ALOHA and PRMA has generally focused on the
performance of these protocols in idealised single cell systems, free from cochannel interference.
Previous studies have normally only considered simple signal propagation models that give little
or no consideration to signal variability. A common assumption made in previous studies
assumes that all packets involved in collisions with other packets are lost, while those packets
received in the absence of other packets are received successfully.
In this thesis, the performance of S-ALOHA and PRMA in more realistic multiple cell
configurations, namely the outdoor cellular and in-building pico-cellular environments, has been
considered. The fluctuations in received signal strengths have been modeled in this thesis using
narrowband statistical propagation models. In addition, the assumption relating to packet
collisions has been reconsidered in this thesis because practical receivers are often capable of
locking onto one packet in the presence of other intra-cell and inter-cell interfering packets, due
to the receiver capture effect. A capture probability model has been developed in this thesis,
which defines the probability of a desired packet capturing the central cell base station in the
presence of both intra-cell and inter-cell interfering packets. Accompanying Markov analysis
theory has been developed to analyse the performance of multiple cell S-ALOHA and PRMA
systems.
Having developed the necessary framework, it has been possible to successfully analyse various
multiple cell S-ALOHA and PRMA systems. These include a multiple cell data-only S-ALOHA
system, multiple cell speech-only PRMA systems with and without selection diversity and
speech packet retransmission and an integrated speech-data multiple cell PRMA system with
selection diversity and speech packet retransmission.