| 1. |
- Corcoran, Diarmuid, et al.
(författare)
-
A Sample Efficient Multi-Agent Approach to Continuous Reinforcement Learning
- 2022
-
Ingår i: Proceedings of the 2022 18th International Conference of Network and Service Management. - : Institute of Electrical and Electronics Engineers Inc.. - 9783903176515 ; , s. 338-344
-
Konferensbidrag (refereegranskat)abstract
- As design, deployment and operation complexity increase in mobile systems, adaptive self-learning techniques have become essential enablers in mitigation and control of the complexity problem. Artificial intelligence and, in particular, reinforcement learning has shown great potential in learning complex tasks through observations. The majority of ongoing reinforcement learning research activities focus on single-Agent problem settings with an assumption of accessibility to a globally observable state and action space. In many real-world settings, such as LTE or 5G, decision making is distributed and there is often only local accessibility to the state space. In such settings, multi-Agent learning may be preferable, with the added challenge of ensuring that all agents collaboratively work towards achieving a common goal. We present a novel cooperative and distributed actor-critic multi-Agent reinforcement learning algorithm. We claim the approach is sample efficient, both in terms of selecting observation samples and in terms of assignment of credit between subsets of collaborating agents.
|
|
| 2. |
- Corcoran, Diarmuid, et al.
(författare)
-
Efficient Real-Time Traffic Generation for 5G RAN
- 2020
-
Ingår i: Proceedings of IEEE/IFIP Network Operations and Management Symposium 2020. - : Institute of Electrical and Electronics Engineers Inc.. - 9781728149738
-
Konferensbidrag (refereegranskat)abstract
- Modern telecommunication and mobile networks are increasingly complex from a resource management perspective, with diverse combinations of software and infrastructure elements that need to be configured and tuned for efficient operation with high quality of service. Increased real-time automation at all levels and time-frames is a critical tool in controlling this complexity. A key component in automation is practical and accurate simulation methods that can be used in live traffic scenarios. This paper introduces a new method with supporting algorithms for sampling key parameters from live or recorded traffic which can be used to generate large volumes of synthetic traffic with very similar rate distributions and temporal characteristics. Multiple spatial renewal processes are used to generate fractional Gaussian noise, which is scaled and transformed into a log-normal rate distribution with discrete arrival events, fitted to the properties observed in given recorded traces. This approach works well for modelling large user aggregates but is especially useful for medium sized and relatively small aggregates, where existing methods struggle to reproduce the most important properties of recorded traces. The technique is demonstrated through experimental comparisons with data collected from an operational LTE network to be highly useful in supporting self-learning and automation algorithms which can ultimately reduce complexity, increase energy efficiency, and reduce total network operation costs.
|
|
| 3. |
- Corcoran, Diarmuid, et al.
(författare)
-
Reinforcement Learning for Automated Energy Efficient Mobile Network Performance Tuning
- 2021
-
Ingår i: Proceedings of the 2021 17th International Conference on Network and Service Management. - : Institute of Electrical and Electronics Engineers (IEEE). ; , s. 216-224, s. 216-224
-
Konferensbidrag (refereegranskat)abstract
- Modern mobile networks are increasingly complex from a resource management perspective, with diverse combinations of software, infrastructure elements and services that need to be configured and tuned for correct and efficient operation. It is well accepted in the communications community that appropriately dimensioned, efficient and reliable configurations of systems like 5G or indeed its predecessor 4G is a massive technical challenge. One promising avenue is the application of machine learning methods to apply a data-driven and continuous learning approach to automated system performance tuning. We demonstrate the effectiveness of policy-gradient reinforcement learning as a way to learn and apply complex interleaving patterns of radio resource block usage in 4G and 5G, in order to automate the reduction of cell edge interference. We show that our method can increase overall spectral efficiency up to 25% and increase the overall system energy efficiency up to 50% in very challenging scenarios by learning how to do more with less system resources. We also introduce a flexible phased and continuous learning approach that can be used to train a bootstrap model in a simulated environment after which the model is transferred to a live system for continuous contextual learning.
|
|
| 4. |
|
|
| 5. |
- Kreuger, Per
(författare)
-
A generative mobility model
- 2021
-
Rapport (övrigt vetenskapligt/konstnärligt)abstract
- Many applications in the mobile radio network domain employ simulations to explore e.g. parameter configurations, robustness of protocols and buffer allocation algorithms. User mobility is (together with traffic and radio propagation models), one of the main components of such simulations, and has large impact on e.g. load distribution, cell handover frequency, signal fading and interference. In many simulations, detailed user mobility is as crucial as the physical infrastructure, where the exact position affect fading and reflections, but in others, e.g. load balancing, handover management and radio resource scheduling, coarser models are often sufficient. But even in these cases, properties of the trajectories of individual users will affect the results of the simulation, e.g. where distribution of positions, rest times, and displacement magnitudes and velocities, need to be considered.
|
|
| 6. |
- Sandbaumhüter, Friederike A., et al.
(författare)
-
Well-Plate muFASP for Proteomic Analysis of Single Pancreatic Islets
- 2022
-
Ingår i: Journal of Proteome Research. - : American Chemical Society (ACS). - 1535-3893 .- 1535-3907. ; 21:4, s. 1167-1174
-
Tidskriftsartikel (refereegranskat)abstract
- Filter-aided sample preparation (FASP) is widely used in bottom-upproteomics for tryptic digestion. However, the sample recovery yield of this methodis limited by the amount of the starting material. While similar to 100 ng of digested protein issufficient for thorough protein identification, proteomic information gets lost with aprotein content <10 mu g due to incomplete peptide recovery from thefilter. Wedeveloped and optimized aflexible well-plate mu FASP device and protocol that issuitable for an similar to 1 mu g protein sample. In 1 mu g of HeLa digest, we identified 1295 +/- 10proteins with mu FASP followed by analysis with liquid chromatography-massspectrometry. In contrast, only 524 +/- 5 proteins were identified with the standardFASP protocol, while 1395 +/- 4 proteins were identified in 20 mu g after standard FASPas a benchmark. Furthermore, we conducted a combined peptidomic and proteomicstudy of single pancreatic islets with well-plate mu FASP. Here, we separated neuropeptides and digested the remaining on-filterproteins for bottom-up proteomic analysis. Our results indicate inter-islet heterogeneity for the expression of proteins involved inglucose catabolism, pancreatic hormone processing, and secreted peptide hormones. We consider our method to provide a usefultool for proteomic characterization of samples where the biological material is scarce. All proteomic data are available under DOI:10.6019/PXD029039
|
|
