Adaptive dynamic capacity borrowing in road-covering mobile networks

This paper introduces adaptive dynamic capacity borrowing strategies for wireless networks covering a road. In a F/TDMA-based model, road traffic prediction models are used to characterise the movement of hot spots, such as traffic jams, and subsequently to predict the teletraffic load offered to the wireless network. A dynamic upper bound on the capacity required to achieve a specified Quality of Service level in the cells is computed. Restricting borrowing to neighbouring cells to avoid excessive re-allocation of capacity, optimal capacity borrowing strategies based on traffic movement and traffic density are given. These strategies can be characterised by a straightforward rule of thumb: borrow capacity from the cell on the steeper side of the traffic peak, that makes our strategy easily implementable. Results of a dynamic simulation under realistic load indicate a significant reduction of call blocking probabilities under our optimal capacity borrowing strategy. A possible extension to CDMA-based systems is briefly indicated.