A typical objective of frequency planning in wireless networks is maximizing the overall network throughput by reducing inter-cell interference. We argue that network throughput alone, as obtained by simulations or direct measurement, is not an adequate measure for evaluating a frequency plan, so that any direct optimization of network throughput would indeed not be advisable as it is especially inadequate for fairness. Contrary to intuition, we demonstrate via simple experiments that in IEEE 802.11 wireless LANs (WLANs) a bad frequency planning, which causes higher interference, may increase the overall network throughput. Such cases occur when far users with low bit-rates get starved under bad frequency planning, while near users that use high bit-rates monopolize the transmission opportunities. Such a counter-intuitive phenomenon indicates that both network throughput and fairness must be considered for frequency planning in WLANs. In this paper we first devise a novel framework to model the load of WLAN cells considering inter-cell interference. Then, we present an efficient frequency planning algorithm which is designed on the basis of this load model. Extensive simulations show that the proposed model allows adequate evaluation of the performance of a frequency plan from both throughput and fairness perspectives. It is also shown that our algorithm provides fairer service to its users compare to existing frequency planning algorithms, while preserving high network utilization.
All Science Journal Classification (ASJC) codes
- Computer Networks and Communications