Performance analysis of the interference adaptation dynamic channel allocation technique in wireless communication networks

Dynamic channel allocation (DCA) problem is one of the major research topics in the wireless networking area. The purpose of this technique is to relieve the contradiction between the increasing traffic load in wireless networks and the limited bandwidth resource across the air interface. The challenge of this problem comes from the following facts: (a) even the basic DCA problem is shown to be NP-complete (none polynomial complete); (b) the size of the state space of the problem is very large; and (c) any practical DCA algorithm should run in real-time.; Many heuristic DCA schemes have been proposed in the literature. It has been shown through simulation results that the interference adaptative dynamic channel allocation (IA-DCA) scheme is a promising strategy in Time Devision Multiple Accesss/Frequency Devision Multiple Accesss (TDMA/FDMA) based wireless communication systems. However, the analytical work on the IA-DCA strategy in the literature is nearly blank.; The performance of a DCA algorithm in TDMA/FDMA wireless systems is influenced by three factors: representation of the interference, traffic fluctuation, and the processing power of the algorithm. The major obstacle in analyzing IA-DCA is the computation of co-channel interference without the constraint of conventional channel reuse factors. To overcome this difficulty, one needs a representation pattern which can approximate the real interference distribution as accurately as desired, and is also computationally viable. For this purpose, a concept called channel reuse zone (CRZ) is introduced and the methodology of computing the area of a CRZ with an arbitrary, non-trivial channel reuse factor if defined. Based on this new concept, the computation of both downlink and uplink co-channel interference is investigated with two different propagation models, namely a simplified deterministic model and a shadowing model. For the factor of the processing power, we proposed an idealized Interference Adaptation Maximum Packing (IAMP) scheme, which gives the upper bound of all IA-DCA schemes in terms of the system capacity. The effect of traffic dynamics is delt with in two steps. First, an asymptotic performance bound for the IA-DCA strategy is derived with the assumption of an arbitrarily large number of channels in the system. Then the performance bound for real wireless systems with the IA-DCA strategy is derived by alleviating this assumption. Our analytical result is compared with the performance bound drawn by Zander and Eriksson for reuse-partitioning DCA{dollar}sp1{dollar} and some simulation results for IA-DCA in the literature. It turns out that the performance bound obtained in this work is much tighter than Zander and Eriksson's bound and is in agreement with simulation results. ftn{dollar}sp1{dollar}only available for deterministic propagation model and downlink connection.