Multi-radio diversity in wireless networks

This paper describes the Multi-Radio Diversity (MRD) wireless system, which uses path diversity to improve loss resilience in wireless local area networks (WLANs). MRD coordinates wireless receptions among multiple radios to improve loss resilience in the face of path-dependent frame corruption over the radio. MRD incorporates two techniques to recover from bit errors and lower the loss rates observed by higher layers, without consuming much extra bandwidth. The first technique is frame combining, in which multiple, possibly erroneous, copies of a given frame are combined together in an attempt to recover the frame without retransmission. The second technique is a low-overhead retransmission scheme called request-for-acknowledgment (RFA), which operates above the link layer and below the network layer to attempt to recover from frame combining failures. We present an analysis that determines how the parameters for these algorithms should be chosen. We have designed and implemented MRD as a fully functional WLAN infrastructure based on 802.11a. We evaluate the MRD system under several different physical configurations, using both UDP and TCP, and measured throughput gains up to 3× over single radio communication schemes employing 802.11’s autorate adaptation scheme. Computer and Communication Sciences, EPFL, Switzerland. Allen Miu received his Ph.D. degree at the Massachusetts Institute of Technology in 2006 and is currently a wireless systems architect at Ruckus Wireless, Inc. He received his S.M. in Computer Science from MIT and a B.Sc. with highest honors in Electrical Engineering and Computer Science from the University of California at Berkeley. He previously worked on the Cricket Indoor Location System and was a research intern at Microsoft Research, Redmond in 2000 and Hewlett-Packard Laboratories, Palo Alto in 2002. His research interests include wireless networks, location systems, mobile computing, and embedded systems. Hari Balakrishnan is an Associate Professor in the EECS Department and a member of the Computer Science and Artificial Intelligence Laboratory (CSAIL) at MIT. His research interests is in the area of networked computer systems. In addition to many widely cited papers, several systems developed as part of his research are available in the public domain. He received a Ph.D. in Computer Science from the University of California at Berkeley in 1998 and a B.Tech. from the Indian Institute of Technology (Madras) in 1993. His honors include an Alfred P. Sloan Research Fellowship (2002), an NSF CAREER Award (2000), the ACM doctoral dissertation award for his work on reliable data transport over wireless networks (1998), and seven award-winning papers at various top conferences and journals, including the IEEE Communication Society’s William R. Bennett Prize (2004). He has also received awards for excellence in teaching and research at MIT (Spira, Junior Bose, and Harold Edgerton faculty achievement awards). C. Emre Koksal received his B.S. degree in Electrical Engineering from the Middle East Technical University, Ankara in 1996. He received his S.M. and Ph.D. degrees from MIT in Electrical Engineering and Computer Science in 1998 and 2002 respectively. He was a postdoctoral fellow in the Networks and Mobile Systems Group in the Computer Science and Artificial Intelligence Laboratory at MIT until 2003. Since then he has been a senior researcher jointly in the Laboratory for Computer Communications and the Laboratory for Information Theory at EPFL, Switzerland. His general areas of interest are wireless communications, computer networks, information theory, stochastic processes and financial economics. He also has a certificate on Financial Technology from the Sloan School of Management at MIT.