Cost-Efficient Allocation of Additional Resources for the Service Placement Problem in Next-Generation Internet

Abstract EN

One of the major challenges in next-generationInternet is to allocate services to nodes in the network.

Thisproblem, known as the service placement problem, can be solvedby layered graph approach.

However, due to the existence ofresource bottleneck, the requests are rejected from the beginningin the resource constrained network.

In this paper we proposetwo iterative algorithms for efficient allocation of additionalresources in order to improve the ratio of accepted serviceplacement requests.

To this end, we (1) introduce a new conceptof sensitivity for each service node to locate the bottlenecknode, (2) state the problem of allocating additional resources,and (3) use sensitivity to propose a simple iterative algorithmand an utilization-based iterative algorithm for efficient resourceallocation.

The performance of these two algorithms is evaluatedby simulation experiments in a variety of parameter settings.

The results show that the proposed algorithms increase requestacceptance ratio significantly by allocating additional resourcesinto the bottleneck node and links.

The utilization-based iterativealgorithm also decreases the long-term cost by making efficientuse of additional resources.