TY - JOUR
T1 - Site-specific knowledge and interference measurement for improving frequency allocations in wireless networks
AU - Chen, Jeremy K.
AU - de Veciana, Gustavo
AU - Rappaport, Theodore S.
N1 - Funding Information:
Manuscript received July 26, 2008. First published December 12, 2008; current version published May 11, 2009. This work was supported by the National Science Foundation under Grant ACI-0305644 and Grant CNS-0325788. This paper was presented in part at the IEEE Vehicular Technology Conference, Baltimore MD, October 2007 [1] and the IEEE Global Communications Conference, Washington, DC, November 2007 [2]. The review of this paper was coordinated by Dr. K. Zangi.
PY - 2009
Y1 - 2009
N2 - We present new frequency allocation schemes for wireless networks and show that they outperform all other published work. Two categories of schemes are presented: 1) those purely based on measurements and 2) those that use site-specific knowledge, which refers to knowledge of building layouts, the locations and electrical properties of access points (APs), users, and physical objects. In our site-specific knowledge-based algorithms, a central network controller communicates with all APs and has site-specific knowledge so that it can a priori predict the received power from any transmitter to any receiver. Optimal frequency assignments are based on predicted powers to minimize interference and maximize throughput. In our measurement-based algorithms, clients periodically report in situ interference measurements to their associated APs; then, the APs' frequency allocations are adjusted based on the reported measurements. Unlike other work, we minimize interference seen by both users and APs, use a physical model rather than a binary model for interference, and mitigate the impact of rogue interference. Our algorithms consistently yield high throughput gains, irrespective of the network topology, AP activity level, number of APs, rogue interferers, and available channels. Our algorithms outperform the best published work by 18.5%, 97.6%, and 1180% for median, 25th percentile, and 15th percentile user throughputs, respectively.
AB - We present new frequency allocation schemes for wireless networks and show that they outperform all other published work. Two categories of schemes are presented: 1) those purely based on measurements and 2) those that use site-specific knowledge, which refers to knowledge of building layouts, the locations and electrical properties of access points (APs), users, and physical objects. In our site-specific knowledge-based algorithms, a central network controller communicates with all APs and has site-specific knowledge so that it can a priori predict the received power from any transmitter to any receiver. Optimal frequency assignments are based on predicted powers to minimize interference and maximize throughput. In our measurement-based algorithms, clients periodically report in situ interference measurements to their associated APs; then, the APs' frequency allocations are adjusted based on the reported measurements. Unlike other work, we minimize interference seen by both users and APs, use a physical model rather than a binary model for interference, and mitigate the impact of rogue interference. Our algorithms consistently yield high throughput gains, irrespective of the network topology, AP activity level, number of APs, rogue interferers, and available channels. Our algorithms outperform the best published work by 18.5%, 97.6%, and 1180% for median, 25th percentile, and 15th percentile user throughputs, respectively.
KW - Cellular networks
KW - Frequency allocation
KW - Radio spectrum management
KW - Site-specific knowledge
KW - Wireless local area network (WLAN)
UR - http://www.scopus.com/inward/record.url?scp=66449110560&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=66449110560&partnerID=8YFLogxK
U2 - 10.1109/TVT.2008.2010862
DO - 10.1109/TVT.2008.2010862
M3 - Article
AN - SCOPUS:66449110560
SN - 0018-9545
VL - 58
SP - 2366
EP - 2377
JO - IEEE Transactions on Vehicular Technology
JF - IEEE Transactions on Vehicular Technology
IS - 5
ER -