TY - GEN
T1 - Estimated-TCP-throughput maximization based routing
AU - Takahashi, H.
AU - Saito, M.
AU - Aida, H.
AU - Tobe, Y.
AU - Tokuda, H.
N1 - Publisher Copyright:
© 2003 IEEE.
Copyright:
Copyright 2015 Elsevier B.V., All rights reserved.
PY - 2003
Y1 - 2003
N2 - This paper presents a new routing scheme called estimated-TCP-throughput maximization based routing (ETR) scheme for mobile ad hoc networks (MANET). Many routing protocols have been proposed to build multi-hop networks in MANET. Most of these protocols select a path based simply on hop metrics, although the shortest path is not always the best path under the influence of radio link quality. We propose that the path should be selected with the maximum end-to-end TCP throughput because most of applications use TCP. In ETR, as a new metric for routing, we introduce an estimated TCP-throughput. A source node measures loss and delay for a path with probing packets periodically. The node estimates maximum TCP throughput from the expression of first approximation with those measured values, and exchanges a path dynamically so that the performance of the using path will be the best of all paths. This paper shows the, design and implementation of ETR. We add ETR to dynamic source routing (DSR) protocol in our MANET testbed, and evaluate ETR. In the case of 50% packet loss at DSR layer. ETR improves TCP throughput by 93% than unmodified DSR. ETR can select the path whose TCP throughput is maximum from the multiple paths to a destination.
AB - This paper presents a new routing scheme called estimated-TCP-throughput maximization based routing (ETR) scheme for mobile ad hoc networks (MANET). Many routing protocols have been proposed to build multi-hop networks in MANET. Most of these protocols select a path based simply on hop metrics, although the shortest path is not always the best path under the influence of radio link quality. We propose that the path should be selected with the maximum end-to-end TCP throughput because most of applications use TCP. In ETR, as a new metric for routing, we introduce an estimated TCP-throughput. A source node measures loss and delay for a path with probing packets periodically. The node estimates maximum TCP throughput from the expression of first approximation with those measured values, and exchanges a path dynamically so that the performance of the using path will be the best of all paths. This paper shows the, design and implementation of ETR. We add ETR to dynamic source routing (DSR) protocol in our MANET testbed, and evaluate ETR. In the case of 50% packet loss at DSR layer. ETR improves TCP throughput by 93% than unmodified DSR. ETR can select the path whose TCP throughput is maximum from the multiple paths to a destination.
KW - Computer networks
KW - Routing
UR - http://www.scopus.com/inward/record.url?scp=34547289840&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=34547289840&partnerID=8YFLogxK
U2 - 10.1109/LCN.2003.1243119
DO - 10.1109/LCN.2003.1243119
M3 - Conference contribution
AN - SCOPUS:34547289840
T3 - Proceedings - Conference on Local Computer Networks, LCN
SP - 120
EP - 129
BT - Proceedings - 28th Annual IEEE International Conference on Local Computer Networks, LCN 2003, Held in conjunction with the Workshop on High-Speed Local Networks, HSLN 2003 and Workshop on Wireless Local Networks, WLN 2003
PB - IEEE Computer Society
T2 - 28th Annual IEEE International Conference on Local Computer Networks, LCN 2003, Held in conjunction with the Workshop on High-Speed Local Networks, HSLN 2003 and Workshop on Wireless Local Networks, WLN 2003
Y2 - 20 October 2003 through 24 October 2003
ER -