A5062212122- Network Traffic and Congestion Control
- Ferroelectric and Negative Capacitance Devices
- Advanced Memory and Neural Computing
- Image and Video Quality Assessment
- Internet Traffic Analysis and Secure E-voting
- Video Coding and Compression Technologies
- Neural Networks Stability and Synchronization
- Advanced Queuing Theory Analysis
- Advanced Wireless Communication Technologies
- Neural dynamics and brain function
- Millimeter-Wave Propagation and Modeling
- Network Security and Intrusion Detection
- Radio Wave Propagation Studies
- Neuroscience and Neural Engineering
- Graph theory and applications
- IoT and Edge/Fog Computing
- Energy Efficient Wireless Sensor Networks
- Interconnection Networks and Systems
- Privacy-Preserving Technologies in Data
- Advanced MIMO Systems Optimization
- CCD and CMOS Imaging Sensors
- Complex Network Analysis Techniques
- Modular Robots and Swarm Intelligence
- Security in Wireless Sensor Networks
- IPv6, Mobility, Handover, Networks, Security
Ericsson (Hungary)
2001-2023
Ericsson (Sweden)
2002-2022
We provide an overview of the most recent advancements and outcomes European 6G flagship project Hexa-X, on topic in-network Artificial Intelligence (AI) Machine Learning (ML). first a general introduction to its ambitions in terms use cases (UCs), key performance indicators (KPIs), value (KVIs). Then, we identify specific challenges realize, implement, enable native integration AI ML 6G, both as means for designing flexible, low-complexity, reconfigurable networks ( <italic...
Detailed knowledge about the traffic mixture is essential for network operators and administrators, as it a key input numerous management activities. Several classification approaches co-exist in literature, but none of them performs well all different application types present Internet. In this study we compare benchmark currently known methods on traces captured an operational 3G mobile network. Utilizing experiences strengths weaknesses existing approaches, novel combined method proposed...
Detailed knowledge about the traffic mixture is essential for network operators and administrators, as it a key input numerous management activities. Several classification approaches co-exist in literature, but none of them performs well all different application traf
We consider the connection between packet loss ratio (PLR) in a switch with finite buffer of size L and tail distribution corresponding infinite queue Q. In literature PLR is often approximated probability P(Q > L), practice latter good conservative estimate on PLR. Therefore, efforts have mainly focused finding bounds asymptotic expressions concerning probabilities queue. However, our first result shows that PLR/P(Q L) can be arbitrary, particular larger than probability. also determine an...
Spiking neural networks (SNN) are expected to enable several use-cases in future communication (beyond 5G and 6G), as edge AI battery-constrained systems can leverage the fast computation high-power efficiency offered by SNNs. In this work we consider a Distributed Wireless SNN (DW-SNN) system analyze its performance terms of inference accuracy total activity when radio losses applied spikes transferred during phase. Our aim is understand how impact considering different spike types, i.e.,...
It is widely accepted that the video frame size process long range dependent (LRD), which has been indicated by a number of statistical tests. This approach suggests use LRD traffic models for engineering, in many cases leads to subexponential queue tails network buffers. Basically, there are two concepts considered as origin modeling: self-similarity and heavy tailed level durations. In this paper we argue new approach, where sequence described three processes representing explicit time...
A significant innovation for future indoor wireless networks is the use of mmWave frequency band. However, an important challenge comes from restricted propagation conditions in this band, which necessitates beamforming and associated beam management procedures, including, instance, tracking or prediction. possible solution to problem artificial-intelligence-based procedures learn hidden spatial patterns channel knowledge predict best directions. In paper, we present a...
One emerging trend in the evolution of sensor technology is to perform Artificial Intelligence (AI) based data processing already on sensor. The benefits are terms reduced upstream communication load or reaction times when applied for example, some time critical control loop. Such developments, however, may easily get conflict with equally important requirement energy efficiency and low power operation. brain inspired neuromorphic resolve this conflict, if we can put all computation tasks...