Efficient Global Optimal Resource Allocation in Interference Networks

Add to your list(s) Download to your calendar using vCal

• Bho Matthiesen, Technische Universität Dresden
• Thursday 09 May 2019, 11:00-12:00
• Room 909B, Electrical and Electronic Engineering.

If you have a question about this talk, please contact Vanessa Rodriguez-Gonzalez.

Resource allocation in interference networks is essential to obtain optimal performance and resource utilization. Unless orthogonal transmission schemes are employed, the optimization problems are non-convex and their solution requires, in general, significant computational resources. Hence, practical systems usually use algorithms with no or only weak optimality guarantees for complexity reasons. Nevertheless, asserting the quality of these algorithms requires the knowledge of the globally optimal solution to these problems. State-of-the-art global optimization approaches mostly employ the monotonic optimization framework which has some major drawbacks, especially when dealing with fractional objectives or complicated feasible sets.

In this talk, two novel global optimization frameworks are presented that allow the optimal solution of previously intractable problems. First, mixed monotonic programming is introduced. Its major benefit over state-of-the-art approaches is a novel bounding technique that allows to directly optimize fractional objectives without requiring Dinkelbach’s algorithm. It also produces tighter bounds that lead to faster convergence. Second, issues with “hard” constraints sets and their solution with the successive incumbent transcending (SIT) scheme are discussed. An resource allocation framework implementing this SIT scheme is presented. Besides having higher numerical stability than the state-of-the-art it also exploits partial convexity properties of the optimization problem. This significantly reduces the computational complexity in problems that are non-convex only due to a few optimization variables.

We apply these novel algorithms to throughput and energy efficiency maximization in wireless interference networks. For a multi-way relay channel with amplify-and-forward relaying we show that joint decoding can outperform treating interference as noise significantly (e.g., 12% on average at a SNR of 10dB) and that computation times for the global optimal solution are several orders of magnitude shorter than with state-of-the-art methods.

Bho Matthiesen received the Diplom-Ingenieur (M.S.) degree in electrical engineering from the Technische Universität Dresden, Germany, in 2012. Since May 2012, he has been a Research Associate with the Chair of Communications Theory at Technische Universität Dresden, where he is currently finishing his Ph.D. degree. His research interests are in the area of signal processing and communication theory, with a focus on global optimization methods for resource allocation.

This talk is part of the Featured talks series.

This talk is included in these lists:

• Andrea Picciau's list
• COMMSP & CP list
• COMMSP Seminar
• Featured talks
• Room 909B, Electrical and Electronic Engineering
• SAP Talks

Note that ex-directory lists are not shown.