In this paper, we propose a novel algorithm for video bitrate adaptation in HTTP Adaptive Streaming (HAS), based on online learning. The proposed algorithm, named Learn2Adapt (L2A) , is shown to provide a robust bitrate adaptation strategy which, unlike most of the state-of-the-art techniques, does not require parameter tuning, channel model assumptions, or application-specific adjustments. These properties make it very suitable for mobile users, who typically experience fast variations in channel characteristics. Experimental results, over real 4G traffic traces, show that L2A improves on the overall Quality of Experience (QoE) and in particular the average streaming bitrate, a result obtained independently of the channel and application scenarios.

We study a class of online convex optimization problems with long-term budget constraints that arise naturally as reliability guarantees or total consumption constraints. In this general setting, prior work by Mannor et al. (2009) has shown that achieving no regret is impossible if the functions defining the agent’s budget are chosen by an adversary. To overcome this obstacle, we refine the agent’s regret metric by introducing the notion of a "K-benchmark", i.e., a comparator which meets the problem’s allotted budget over any window of length K. The impossibility analysis of Mannor et al. (2009) is recovered when K=T; however, for K=o(T), we show that it is possible to minimize regret while still meeting the problem’s long-term budget constraints. We achieve this via an online learning policy based on Cautious Online Lagrangiant Descent (COLD) for which we derive explicit bounds, in terms of both the incurred regret and the residual budget violations.

This paper addresses a fundamental challenge in cloud computing, that of learning an economical yet robust reservation, i.e. reserve just enough resources to avoid both violations and expensive over provisioning. Prediction tools are often inadequate due to observed high variability in CPU and memory workload. We propose a novel model-free approach that has its root in online learning. Specifically, we allow the workload profile to be engineered by an adversary who aims to harm our decisions, and we investigate a class of policies that aim to minimize regret (minimize losses with respect to a baseline static policy that knows the workload sample path). Then we propose a combination of the Lyapunov optimization theory [1] and a linear prediction of the future based on the recent past, used in learning and online optimization problems, see [2]. This enables us to come up with a no regret policy, i.e., a policy whose cost difference to the benchmark and violation constraint residual both grow sublinearly in time, and hence become amortized over the horizon. Our policy has then “no regret and eventually learns the minimum cost reservation subject to a time-average constraint for violations.

We study the user association problem in the context of dense networks, where standard adaptive algorithms become ineffective. This paper proposes a novel data-driven technique leveraging the theory of robust optimization. The main idea is to predict future traffic fluctuations, and use the predictions to design association maps before the actual arrival of traffic. Although, the actual playout of the map is random due to prediction error, the maps are robustly designed to handle uncertainty, preventing constraint violations, and maximizing the expectation of a convex utility function, which is used to accurately balance base station loads. We propose a generalized iterative algorithm, referred to as GRMA, which is shown to converge to the optimal robust map. The optimal maps have the intriguing property that they jointly optimize the predicted load and the variance of the prediction error. We validate our robust maps in Milano-area traces, with dense coverage and find that we can reduce violations from 25% (inflicted by a baseline adaptive algorithm) down to almost zero.

We study the problem of user association in Ultra Dense Networks (UDNs) for two network services; one requiring QoS guarantees to VIP flows, and one best effort service. The goal is to take advantage of statistical multiplexing in order to optimize the use of resources, while ensuring that the VIP flows enjoy active performance guarantees. We formulate this as an optimization problem, show that the problem is convex, and finally demonstrate that the optimum point can in fact be realized by distributed user association rules. In this way, our framework uses the fundamental problem of user association to show that both isolation and statistical multiplexing can be achieved in the context of UDNs, when different services or slices must share BS resources, as envisioned in 5G networks. We demonstrate no violations of the VIP flow constraint on real data traces for mobile network traffic, while a baseline best effort distributed policy applied to this setup inflicts up to 46.5% violations.

The massive scale of future wireless networks will cause computational bottlenecks in performance optimization. In this paper, we study the problem of connecting mobile traffic to Cloud RAN (C-RAN) stations. To balance station load, we steer the traffic by designing device association rules. The baseline association rule connects each device to the station with the strongest signal, which does not account for interference or traffic hot spots, and leads to load imbalances and performance deterioration. Instead, we can formulate an optimization problem to decide centrally the best association rule at each time instance. However, in practice this optimization has such high dimensions, that even linear programming solvers fail to solve. To address the challenge of massive connectivity, we propose an approach based on the theory of optimal transport, which studies the economical transfer of probability between two distributions. Our proposed methodology can further inspire scalable algorithms for massive optimization problems in wireless networks.

We study the user association problem in the context of dense networks, where standard adaptive algorithms become ineffective. The paper proposes a novel data-driven technique leveraging the theory of robust optimization. The main idea is to predict future traffic fluctuations, and use the predictions to design association maps before the actual arrival of traffic. Although the actual playout of the map is random due to prediction error, the maps are robustly designed to handle uncertainty, preventing constraint violations, and maximizing the expectation of a convex utility function, which allows to accurately balance base station loads. We propose a generic iterative algorithm, referred to as GRMA, which is shown to converge to the optimal robust map. The optimal maps have the intriguing property that they jointly optimize the predicted load and the variance of the prediction error. We validate our robust maps in Milano-area traces, with dense coverage and find that we can reduce violations from 25% (achieved by an adaptive algorithm) down to almost zero.