High Capacity CDMA and Collaborative Techniques

Shakya, Indu Lal (2008) High Capacity CDMA and Collaborative Techniques. Doctoral thesis (DPhil), University of Sussex.

[img]
Preview
PDF - Published Version
Download (7MB) | Preview

Abstract

The thesis investigates new approaches to increase the user capacity and improve the error
performance of Code Division Multiple Access (CDMA) by employing adaptive interference cancellation
and collaborative spreading and space diversity techniques. Collaborative Coding Multiple
Access (CCMA) is also investigated as a separate technique and combined with CDMA. The
advantages and shortcomings of CDMA and CCMA are analysed and new techniques for both the
uplink and downlink are proposed and evaluated.
Multiple access interference (MAI) problem in the uplink of CDMA is investigated first. The
practical issues of multiuser detection (MUD) techniques are reviewed and a novel blind adaptive
approach to interference cancellation (IC) is proposed. It exploits the constant modulus (CM)
property of digital signals to blindly suppress interference during the despreading process and obtain
amplitude estimation with minimum mean squared error for use in cancellation stages. Two
new blind adaptive receiver designs employing successive and parallel interference cancellation
architectures using the CM algorithm (CMA) referred to as ‘CMA-SIC’ and ‘BA-PIC’, respectively,
are presented. These techniques have shown to offer near single user performance for large
number of users. It is shown to increase the user capacity by approximately two fold compared
with conventional IC receivers. The spectral efficiency analysis of the techniques based on output
signal-to interference-and-noise ratio (SINR) also shows significant gain in data rate. Furthermore,
an effective and low complexity blind adaptive subcarrier combining (BASC) technique using a
simple gradient descent based algorithm is proposed for Multicarrier-CDMA. It suppresses MAI
without any knowledge of channel amplitudes and allows large number of users compared with
equal gain and maximum ratio combining techniques normally used in practice.
New user collaborative schemes are proposed and analysed theoretically and by simulations
in different channel conditions to achieve spatial diversity for uplink of CCMA and CDMA. First,
a simple transmitter diversity and its equivalent user collaborative diversity techniques for CCMA
are designed and analysed. Next, a new user collaborative scheme with successive interference
cancellation for uplink of CDMA referred to as collaborative SIC (C-SIC) is investigated to reduce
MAI and achieve improved diversity. To further improve the performance of C-SIC under high
system loading conditions, Collaborative Blind Adaptive SIC (C-BASIC) scheme is proposed.
It is shown to minimize the residual MAI, leading to improved user capacity and a more robust
system. It is known that collaborative diversity schemes incur loss in throughput due to the need of
orthogonal time/frequency slots for relaying source’s data. To address this problem, finally a novel
near-unity-rate scheme also referred to as bandwidth efficient collaborative diversity (BECD) is proposed and evaluated for CDMA. Under this scheme, pairs of users share a single spreading sequence to exchange and forward their data employing a simple superposition or space-time
encoding methods. At the receiver collaborative joint detection is performed to separate each
paired users’ data. It is shown that the scheme can achieve full diversity gain at no extra bandwidth
as inter-user channel SNR becomes high.
A novel approach of ‘User Collaboration’ is introduced to increase the user capacity of CDMA
for both the downlink and uplink. First, collaborative group spreading technique for the downlink
of overloaded CDMA system is introduced. It allows the sharing of the same single spreading
sequence for more than one user belonging to the same group. This technique is referred to as
Collaborative Spreading CDMA downlink (CS-CDMA-DL). In this technique T-user collaborative
coding is used for each group to form a composite codeword signal of the users and then a
single orthogonal sequence is used for the group. At each user’s receiver, decoding of composite
codeword is carried out to extract the user’s own information while maintaining a high SINR performance.
To improve the bit error performance of CS-CDMA-DL in Rayleigh fading conditions,
Collaborative Space-time Spreading (C-STS) technique is proposed by combining the collaborative
coding multiple access and space-time coding principles. A new scheme for uplink of CDMA
using the ‘User Collaboration’ approach, referred to as CS-CDMA-UL is presented next. When
users’ channels are independent (uncorrelated), significantly higher user capacity can be achieved
by grouping multiple users to share the same spreading sequence and performing MUD on per
group basis followed by a low complexity ML decoding at the receiver. This approach has shown
to support much higher number of users than the available sequences while also maintaining the
low receiver complexity. For improved performance under highly correlated channel conditions,
T-user collaborative coding is also investigated within the CS-CDMA-UL system.

Item Type: Thesis (Doctoral)
Schools and Departments: School of Engineering and Informatics > Engineering and Design
Subjects: T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5101 Telecommunication Including telegraphy, telephone, radio, radar, television
Depositing User: Indu Lal Shakya
Date Deposited: 09 May 2012 09:13
Last Modified: 09 May 2012 09:13
URI: http://sro.sussex.ac.uk/id/eprint/7619

View download statistics for this item

📧 Request an update