Fast and Cost-Efficient Virtualized Network Function Placement Algorithm in Wireless Multi-hop Networks

Jahedi, Zahra; Kunz, Thomas

doi:10.1007/978-3-030-61746-2_3

Cited by 1 publication

(2 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We measured the execution time, acceptance rate, and placement cost and compared it with mathematical methods and similar heuristics to identify the heuristic that can accept the highest ratio of the requests with lowest complexity and execution time. As mentioned earlier, the cost in our mathematical models and heuristics is calculated the same as the objective function mentioned in Equation ( 2), for wired networks and Equation (10) for wireless networks in Chapter 3. The cost is based on the cost of consumed resources by the VNFs in the physical network that includes the cost of total units of nodal resources used by VNFs, and the cost of total units of bandwidth used by virtual links in the physical network.…”

Section: Results and Analysismentioning

confidence: 99%

“…Our third publication describes the FACE heuristic mentioned in Chapter 4, Section 4.4, and was presented in the 19th International Conference on Ad Hoc Networks and Wireless (AdHoc-Now 2020), titled "Fast and cost-efficient virtualized network function placement algorithm in wireless multi-hop networks" [10]. Our forth publication, "The value of simple heuristics for virtualized network function placement" [11] describes four increasingly complex heuristics: random placement, shortest path placement, all shortest path placement, FACE heuristic, and their performance comparison that are described in Chapter 4.…”

Section: Publicationsmentioning

confidence: 99%

See 1 more Smart Citation

Simple and Scalable Approach for Virtualized Network Function Placement in Wireless Multi-hop Networks

Jahedi¹

Self Cite

View full text Add to dashboard Cite

Network Function Virtualization (NFV) can lower the CAPEX and/or OPEX for service providers and allow for quick deployment of services. The main challenge in the use of Virtualized Network Functions (VNF) is the VNFs' placement in the network. This research provides mathematical models and heuristics for NF placement for wired and wireless networks. We use Integer Linear and Non-Linear Programming as a mathematical optimization program for NF placement. We start from a basic model for a wired network and extend it gradually to develop a traffic-aware mathematical model for NF placement in wireless multi-hop networks. For the first time, we model the interference which is a major difference between a wired and wireless network and included it in our optimization model. We identified the issue of scarcity of BW in wireless multi-hop networks and its role in the average cost of placement and acceptance rate of requests. The critical problem of mathematical models is that they are NP-hard, and consequently not applicable to larger networks. While there exist many efforts in designing a heuristic model that can provide solutions in a timely manner, the primary focus with such heuristics was almost always whether they provide near-optimal results. Consequently, the heuristics themselves become quite non-trivial, and solving the placement problem for larger networks still takes a significant amount of time. In our research, in contrast, we focus on designing a simple and scalable heuristic. We propose a set of heuristics, which are gradually becoming more complex. We start from the random placement heuristic as the simplest approach and at each step add a parameter such as choosing between shortest paths, sort NFs based on their nodal resources, and replacing previously placed NFs to our heuristic. We compare the performance of our heuristics with each other, related heuristics, and our mathematical model. Our results demonstrate that the simple approach of placing NFs along their shortest path can find near-optimal solutions much faster than the other more complicated heuristics while keeping the ratio of accepted requests close to the acceptance ratio of a NP-hard optimization model.Firstly, I would like to express my sincere gratitude to my advisor, Professor Thomas Kunz, for the continuous support of my Ph.D. study and related research, for your patience, motivation, and immense knowledge. Your guidance helped me in all the time of research and writing of this thesis. Your constant willingness to share ideas with me regarding my research, in spite of your busy schedule, is sincerely appreciated.I cannot express my unfailing gratitude and love to my caring, loving husband, Arash Karami-Mohammadi, who has supported me throughout this process. Your encouragement when the times got rough is much appreciated.Last, but not least, I owe the greatest debt to my parents for not only their neverending affection, love, and patience, but also their encouragement and unconditional support to me in all my decisions. Their guidanc...

show abstract

Section: Results and Analysismentioning

confidence: 99%

Section: Publicationsmentioning

confidence: 99%