An optimized bandwidth allocation in broadband digital economy

Efficient and optimized data transfer in broadband is the anchor point of digital economy in modern era. One the most challenging issues for providing different services in broadband is the bandwidth allocation. That is a real-time multi dimensional problem solving requiring to satisfy QOS and SLA constraints at the same time. In this paper, we present a real-time adaptive framework for the problem of optimal bandwidth allocation to broadband services in Iran. Our approach for the allocation is based on Network Utility Maximization. In this way, the optimal bandwidth is the one which maximizes network users' utility. For this purpose, the bandwidth is dynamically allocated to the services with highest importance for the user (according to the SLA) and the remaining bandwidth is most appropriately allocated to the less important services. Each service has its own (non-linear) utility function which is considered in each allocation cycle. The utility values are calculated in real-time and the bandwidth allocation pattern which maximizes the total users' utilities is selected. Simulation results showed satisfactory bandwidth allocation patterns in different scenarios according to the network utility and user preferences in SLA.

I. Introduction

The digital economy is developing rapidly worldwide. It is the single most important driver of innovation, competitiveness and growth, and it holds huge potential for entrepreneurs and small and medium-sized enterprises (SMEs) [1]. As data is the key point of digital economy, transferring large amount of them at highest possible speed is crucial. This is why policy makers set certain goals for broadband availability (such as those in the Digital Agenda for Europe) [2] and encourage investment in broadband infrastructure.