Dynamic multi-channel assignment using network flows in wireless data networks

Abstract

The radio frequency spectrum, a scarce resource in mobile communications, has to be efficiently utilized with the objectives of increasing the network capacity and minimizing the interference. A variety of channel assignment strategies have been developed to achieve these objectives. As the cell sizes get smaller, there is a greater need for efficient channel assignment algorithms which are desired to dynamically balance the load of the system. We propose a dynamic multi-channel assignment (DMCA) algorithm where the assignment decision is assisted by the mobiles. Our algorithm is based on the concept of network flows and handles all the events in the system gracefully. Several existing channel assignment algorithms can be easily modeled in the proposed network flow framework model. Simulation results show that DMCA algorithm performs well under heavy traffic conditions and handles different traffic classes gracefully.

Introduction

The growing demand for mobile communications and portable computing devices coupled with the limited allowed radio frequency spectrum for this use, led to the problem of efficient management of resources like bandwidth in wireless networks. Thus evolved the concept of a cellular architecture, in which frequency can be reused in cells at a safe distance apart so as to guarantee no or minimal interference among the cells involved. Existing cellular systems generally use some variations of a fixed channel assignment (FCA) scheme in which a cell is assigned a fixed subset of the available channels [3]. The main drawback of such an assignment is that the system cannot adapt to changes in the channel demands of the user traffic. In some variants of the FCA scheme, a cell can borrow channels from its neighbors to satisfy the additional requests. Also, channels can be migrated from cells which are not necessarily the neighbors of the needy cell [2]. In another class of schemes, called the dynamic channel assignment (DCA) schemes, all the channels are available for assignment to every cell but the usage of these channels is constrained by the interference experienced. Since the reuse of channels is time-varying, the system dynamically attempts to minimize the mutual interference among all active channels [4].

Although the DCA strategies are more flexible and increase the utilization of the channels, yet the high cost of managing the channel allocation and maintenance of interference information becomes a critical issue. For different algorithms on DCA, refer to Ref. [7].

This paper extends the network flow framework proposed in Refs. [8], [9], to design dynamic multi-channel assignment (DMCA) algorithm for supporting wireless multimedia services. The proposed on-line algorithm dynamically assigns channels to users upon demand, and also reassigns channels among users when the interference degrades below a threshold. The assignment decision of a number of channels to a new call is made locally which may possibly involve reassignment of a certain number of active calls to other channels. Reconfiguration is also necessary when a mobile has to be reassigned to a better quality channel (which experiences less interference), if possible. Our algorithm handles reassignments in an incremental manner rather than a global re-computation of assignments.

Our results show that DMCA algorithm handles different traffic classes better than FCA and provides good performance under heavy traffic load conditions.

This paper is organized as follows. Section 2 introduces the network flow framework and the construction of a flow network for DMCA. Section 3 presents the algorithm and explains how the multi-channel assignment, call termination and handoff are handled. This section also presents how the DMCA algorithm attempts to maintain the quality of service for different traffic classes. In Section 4, other channel assignment schemes and their relation to the network flow framework are discussed. Section 5 describes the simulation environment and the experiments performed. Conclusions are given in Section 6. An explanation of the signal propagation model used in the simulations is presented in Appendix.