Search

Federated learning is an effective method to train a machine learning model without requiring to aggregate the potentially sensitive data of agents in a central server. However, the limited communication bandwidth, the hardware of the agents and a potential application-specific latency requirement impact how many and which agents can participate in the learning process at each communication round. In this paper, we propose a selection metric characterizing each agent’s importance with respect to both the learning process and the resource efficiency of its wireless communication channel. Leveraging this importance metric, we formulate a general agent selection optimization problem, which can be adapted to different environments with latency or resource-oriented constraints. Considering an example wireless environment with latency constraints, the agent selection problem reduces to the 0/1 Knapsack problem, which we solve with a fully polynomial approximation. We then evaluate the agent selection policy in different scenarios, using extensive simulations for an example task of object classification of European traffic signs. The results indicate that agent selection policies which consider both learning and channel aspects provide benefits in terms of the attainable global model accuracy and/or the time needed to achieve a targeted accuracy level. However, in scenarios where agents have a limited number of data samples or where the latency requirement is very stringent, a pure learning-based agent selection policy is shown to be more beneficial during the early or late stages of the learning process.

Wireless communication networks provide a critical infrastructure, particularly in emergency situations due to disruptive events such as natural disasters or terrorist attacks. However, in these kinds of scenarios part of the network may no longer be operational and a traffic hotspot may emerge, which may result in coverage and/or capacity issues. Deploying self-steering drone-mounted base stations offers a potential method to quickly restore coverage and/or provide capacity relief in such situations, but appropriate positioning is crucial in order for a drone base station to be truly effective. Motivated by that challenge, we propose a data-driven algorithm to optimize the position of a drone base station in a scenario with a site failure and emergence of a traffic hotspot. We demonstrate that the use of a drone, when properly positioned, yields significant performance gains, and that our algorithm outperforms benchmark mechanisms in a wide range of scenarios. In addition, we show that our algorithm is able to find a near-optimal position for the drone in a reasonable amount of time, and even has the ability to track the optimal position in case of a moving hotspot.

We assess the system-level performance of non-orthogonal spectrum sharing achieved via maximum sum-rate (SR), Nash bargaining (NB), and zero-forcing (ZF) transmit beamforming techniques. A look-up table based physical layer abstraction and radio resource management mechanisms (including packet scheduling) are proposed and incorporated in system-level simulations, jointly with other important aspects of network operation. In the simulated scenarios, the results show similar system-level performance of SR (or NB) as ZF in the context of spectrum sharing, when combined with maximum sum-rate (or proportional fair) packet scheduler. Further sensitivity analysis also shows similar behavior of all three beamforming techniques with regard to the impact on system-level performance of neighbor-cell activity level and feedback error. A more important observation from our results is that, under ideal conditions, the performance enhancement of non-orthogonal spectrum sharing over orthogonal spectrum sharing and fixed spectrum assignment is significant.

Motivated by the "beyond 3g" vision of radio access network integration and coordinated radio resource management, a purely analytical performance assessment is presented for a single access point integrating multiple radio accesses. Principal focus is placed on the evaluation of multi-user diversity, multi-access diversity and trunking gains. Scenarios with persistent and non-persistent data flows are investigated, concentrating on throughput and transfer time performance, respectively. A number of numerical experiments are included in order to quantify the relative contribution of the distinguished aspects to the performance gain. These experiments indicate that the exploitation of multi-user diversity with a channel-aware access selection scheme attains the most significant gains, while also the trunking gain that is due to an above-proportional performance enhancement when aggregating system-specific capacities, is noted to be significant. The assignment of multiple accesses to a given flow is demonstrated to have limited potential., QC 20110218

In ‘Beyond 3G’ wireless networking scenarios, characterized by the cooperation of multiple access networks comprising different radio access technologies, Multi-Radio Resource Management (MRRM) mechanisms fulfill a key role in enhancing system capacity, resource efficiency, coverage and service quality. This overview paper addresses the embedding of MRRM in the Ambient Networks concept, describes the principal MRRM functions and discusses distribution aspects and initial performance evaluation results. This work has been performed as part of the Ambient Networks project, an integrated project within the EU IST 6th Framework Programme. In ‘Beyond 3G’ wireless networking scenarios, characterized by the cooperation of multiple access networks comprising different radio access technologies, Multi-Radio Resource Management (MRRM) mechanisms fulfill a key role in enhancing system capacity, resource efficiency, coverage and service quality. This overview paper addresses the embedding of MRRM in the Ambient Networks concept, describes the principal MRRM functions and discusses distribution aspects and initial performance evaluation results. This work has been performed as part of the Ambient Networks project, an integrated project within the EU IST 6th Framework Programme., QC 20110111