| Evaluating and Improving TCP with Adaptive Transmission and Routing Protocols in Frequency-Hop Wireless Ad Hoc Networks
Abstract
Frequency-hop spread spectrum radios are known useful for military communications due to their effective anti-jamming capability. In a harsh radio environment with dynamic interference and jamming, a set of adaptive protocols was previously proposed by Pursley and colleagues to adapt the radio transmission power, forward error correction code rate, and ad hoc forwarding routes according to the channel conditions. Based on error and erasure information acquired from each frequency-hop transmission, the protocols were able to determine when the channel is less interfered/jammed such that a high (data) rate code and a lower transmit power level can be used, and when it is more interfered such that a low rate code and higher transmit power levels can be used. Multihop routing in such a network is guided by a link state protocol utilizing both the error-erasure metric and an end-to-end transmission energy metric.
With constant rate packet communication among random peers in an ad hoc network, the original study showed that the adaptive protocols attain an excellent balance between network throughput and energy efficiency. The study, however, has not accounted for possible reactive behaviors of an upper layer rate controlled transport protocol, such as TCP, due to the resulting fluctuations in link rates, packet success probabilities, and round-trip latencies. In this paper, through extensive simulations, we study the TCP throughput performance in a frequency-hop network with adaptive transmission and routing protocols. In different network topologies and jamming scenarios, throughputs of TCP-Reno flows are evaluated with and without adaptive protocols. Results show that while adaptive protocols demonstrate substantial energy efficiency, their throughput performance is degraded in most cases. An in-depth analysis of the protocol interaction reveals an apparent trade-off between power saving and throughput enhancement, and such degradations are mostly observed due to overly sensitive retraction in transmission power. An adaptive approach to reduce such sensitivity is incorporated and the resulting throughput improvements are confirmed via simulation results.
|
| Yu Zhou received his B.S. degree in electronics and communication engineering from Sun Yat-sen University, Guangzhou, China in 2002, and the M.S. degree in electrical and computer engineering from Clemson University, Clemson, SC in 2006. He was a Research Assistant in the Wireless Communication Program. His research interests are wireless ad hoc networks, quality of service in wireless networks, and adaptive protocols for packet radio network.
Dr. Kuang-Ching Wang received his Ph.D. degree in electrical and computer engineering from the University of Wisconsin, Madison in 2003. Dr. Wang is currently an Assistant Professor with the Department of Electrical and Computer Engineering at Clemson University, Clemson, South Carolina. From 2000 to 2003, he participated in the DARPA Sensor Information Technology (SensIT) Program and developed location-centric network protocols for collaborative signal processing applications. His research addresses distributed applications and networking methods in wireless sensor networks, quality of service in wireless networks, and experimental research testbeds.
|