Changes between Initial Version and Version 1 of EverettMSThesis

Timestamp:

Feb 17, 2013, 6:02:28 PM (11 years ago)

Author:

evaneverett

Comment:

--

EverettMSThesis

@@ +1 @@
+= Full-Duplex Infrastructure Nodes: Achieving Long-Range with Half-Duplex Mobiles = 
+
+Evan Everett[[BR]]
+MS Thesis[[BR]]
+Rice University Department of Electrical and Computer Engineering[[BR]]
+Defended & Submitted April 2012
+
+== Abstract ==
+One of the primary sources of inefficiency in today's wireless networks is the half-duplex constraint -- the assumption that nodes cannot transmit and receive simultaneously in the same band. The reason for this constraint and the hurdle to full-duplex operation is self-interference: a node's transmit signal appears at its own receiver with very high power, desensitizing the receiver electronics and precluding the reception of a packet from a distant node. Recent research has demonstrated that full-duplex can indeed be feasible by employing a combination of analog and digital self-interference cancellation mechanisms. However, two glaring limitations remain. The first is that the full-duplex state-of-the-art requires at least two antennas and extra RF resources that space-constrained mobile devices may not be able to accommodate. The second limitation is range: current full-duplex demonstrations have been for ranges less than 10 m. At longer distances nodes must transmit with higher power to overcome path loss, and the power differential between the self-interference and the signal-of-interest becomes more that the current cancellation mechanisms can handle.
+
+We therefore present engineering solutions for answering the following driving questions: (a) can we leverage full-duplex in a network consisting mostly of half-duplex mobiles? and (b) can we extend the range of full-duplex by achieving self-interference suppression sufficient for full-duplex to outperform half-duplex at ranges exceeding 100 m?
+
+In answer to the first question, we propose moving the burden of full-duplexing solely to access points (APs), enabling the AP to boost network throughput by receiving an uplink signal from one half-duplex mobile, while simultaneously transmitting a downlink signal to another half-duplex mobile in the same band. In answer to the second question we  propose an AP antenna architecture that uses a careful combination of three mechanisms for passive suppression of self-interference: directional isolation, absorptive shielding, and cross-polarization. Results from a 20 MHz OFDM prototype demonstrate that the proposed AP architecture can achieve 90+ dB total self-interference suppression, enabling 50% uplink rate gains over half-duplex for ranges up to 150 m. 
+
+
+== Thesis ==
+[attachment:wiki:HunterPhDThesis/Files:HunterThesis.pdf?format=raw Distributed Protocols for Signal-Scale Cooperation] (8MB PDF)
+
+== Citation ==
+
+{{{
+
+@phdthesis{HunterPhD,
+  Author = {Christopher Hunter},
+  School = {Rice University},
+  Title = {Distributed Protocols for Signal-Scale Cooperation},
+  Year = {2012},
+  URL = {http://warp.rice.edu/trac/wiki/HunterPhDThesis}}
+}}}