Dissertation Defense, Xingkai Bao

Dissertation Defense, Xingkai Bao


Title:                      Wireless User Cooperation through different forwarding methods, coding
                             design, and for  different applications

Date:                     08/12/2009 
Time:                     10:00-12:00 am
Location:              PA416

Abstract:

In modern wireless systems, the key challenge is multipath fading which causes fluctuations in the received signal power. Cooperative networks, or called user cooperations, is an emerging technique to combat multipath fading by exploiting spatial diversity, where distributed wireless nodes help each other transmit data or conduct signal processing. Since the impairment of the multipath fading can
be largely reduced by cooperative networks, more reliable communications, higher data rate, larger coverage range, and more accurate position localization can be achieved.

The first contribution of this dissertation involves the , an efficient data forwarding scheme which combines the merits of and . Theoretical analysis for the capacity bound are conducted and practical simulation based on turbo codes and LDPC codes are performed. Both of them show that DAF can effectively improve the data rate and outage probability, when the relay node can not decode the data from source node successfully. The second contribution focuses on an efficient space-time-coding design for three-node user cooperation, called . The basic idea is to separate the space-time coding into AF part and DF part, which are transmitted through broadcasting phase and relaying phase respectively. The AF part brings additional space-time diversity before the two users decode each other's data correctly, while the DF part in relaying phase obtains excellent space-time coding performance. The third contribution of this thesis is motivated by recent cooperative techniques through network coding. At first, the framework of is proposed, which dynamically encodes all data packets from all distributed users into a large network codes under the consideration that the inter-user channels are in random outage. The key idea is to match dynamical network-on-graph with codes-on-graph. Then, the joint channel and network coded cooperation, aka, generalized adaptive network coded cooperation (GANCC) is suggested. GANCC integrates channel codes as a seemless part of network codes by a simple operation of circulant interleaving. The fourth contribution of this thesis focuses on an efficient rateless erasure codes, called , which can be used for wireless network coding. It exploits a different way to construct rateless codes, which optimizes both the left and right degree distribution progressively layer by layer. Both theoretical analysis and computer simulation shows that DDD codes achieve better bandwidth efficiency than existing rateless codes.

The fifth contribution addresses the application of cooperative network on position localizations. A distributed space-time coding localization are suggested, where different reference nodes transmit simultaneously using distributed space-time block codes, and each target node estimates the using the synchronization and decoding techniques for space-time codes. This scheme combats the multipath fading and help distinguish the line-of-sight path, thus significantly enhances the localization accuracy.

 

PhD committee:

Prof Tiffany Jing Li, ECE (Committee Chair)
Prof Shalinee Kishore, ECE
Prof Mooi Choo Chuah, CSE
Dr. Kamran Azadet, LSI inc