To date, commercial industrial robots only provide multi-robot coordination for their own fleet of robots and treat robots from other vendors as general obstacles. The ability to enable robots from different vendors to co-exist in the same space is crucial to prevent vendor lock-in. We present the first decentralized system that achieves coordination between a heterogeneous fleet of black box robots for which the internals of the navigation stack are presumed unmodifiable. Our system, which we call CODAK, achieves the coordination by relying on minimum set of interfaces that are commonly available on most industrial and service robots. For each robot, CODAK uses a trained recurrent neural network to anticipate collisions from externally observable metrics. Anticipated collisions are avoided using a simple, but yet effective, concurrency control scheme. We run a series of experiments in simulation and with real robots to demonstrate CODAK’s ability to enable safe navigation in different environments. We also experimentally compare CODAK with previously published white-box solutions to evaluate the penalty of black-box constraint.

Network slicing is a key technique in 5G and beyond for efficiently supporting diverse services. Many network slicing solutions rely on deep learning to manage complex and high-dimensional resource allocation problems. However, deep learning models suffer limited generalization and adaptability to dynamic slicing configurations. In this paper, we propose a novel frame-work that integrates constrained optimization methods and deep learning models, resulting in strong generalization and superior approximation capability. Based on the proposed framework, we design a new neural-assisted algorithm to allocate radio resources to slices to maximize the network utility under inter-slice resource constraints. The algorithm exhibits high scalability, accommodating varying numbers of slices and slice configurations with ease. We implement the proposed solution in a system-level network simulator and evaluate its performance extensively by comparing it to state-of-the-art solutions including deep reinforcement learning approaches. The numerical results show that our solution obtains near-optimal quality-of-service satisfaction and promising generalization performance under different network slicing scenarios.

Network slicing enables operators to virtually partition network resources and instantiate different virtual networks, supporting flexible quality of service, in the form of service level agreements (SLAs). Optimizing resource allocation for network slicing is a complex task, given the dynamicity and randomness of network conditions. Leveraging principles of reinforcement learning, we propose an algorithmic solution to optimize the SLA success rate across a radio access network. Tailoring the algorithm to the problem at hand, and leveraging some prior knowledge on the system to be optimized, we ensure fast convergence, data efficiency, and safe exploration. The solution is scalable both in the number of cells and in the number of slices. Extensive numerical evaluations on a simulated environment show the effectiveness of the proposed solution and its advantages versus both simple baselines and sophisticated solutions based on deep reinforcement learning, in terms of speed of convergence and SLA success rate.

Wireless networks are becoming more and more complex. New radio access technologies and cell densification have provided increased efficiency and met the needs of an increasing traffic demand. On the other hand, such increment in complexity poses new challenges to legacy network management and optimization solutions. Traditionally, network optimization has been performed by applying static, rule-based schemes. Both in academia and industry, there is a consensus about the need to transit from static optimization approaches to more automated, dynamic, cognitive techniques. In this paper we embrace this perspective and we propose a set of cognitive algorithms for network performance optimization. We leverage the electric antenna tilt to achieve improved network performance. The proposed algorithms rely on reinforcement learning principles and are validated in a simulated environment.

In the framework of queuing theory with multiclass jobs, trunk-reservation is an admission control technique to handle job class priority in an online fashion, and serve as many high priority jobs as possible. This is achieved by rejecting jobs with sufficiently low priority when the buffer space becomes a scarce resource, i.e., when the queue may soon overflow. Mathematically, the objective is to maximize the long-term reward of the admitted jobs, where each job is assigned a reward which is monotonic with respect to the priority class it belongs to. In this paper we study online learning of optimal trunk-reservation policies when the system parameters, i.e., class arrival rates and service time, are unknown. Starting from a Markov Decision Process (MDP) formulation, we leverage the stairway structure of the optimal policy to define Integer Gradient Ascent (IGA), a reinforcement learning (RL) algorithm based on policy-gradient methods, specifically tailored to the mathematical properties of the problem at hand. We provide theoretical results on the convergence properties of IGA. Extensive numerical experiments characterize its behavior and confirm that it outperforms standard RL techniques in terms of convergence rate.

To optimize routing of flows in datacenters, SDN controllers receive a packet-in message whenever a new flow appears in the network. Unfortunately, flow arrival rates can peak to millions per second, impairing the ability of controllers to treat them on time. Flow scheduling copes with such sheer numbers by segmenting the traffic between elephant and mice flows and by treating elephant flows in priority, as they disrupt short lived TCP flows and create bottlenecks. We propose a learning algorithm called SOFIA and able to perform optimal online flow segmentation. Our solution, based on stochastic approximation techniques, is implemented at the switch level and updated by the controller, with minimal signaling over the control channel. SOFIA is blind, i.e., it is oblivious to the flow size distribution. It is also adaptive, since it can track traffic variations over time. We prove its convergence properties and its message complexity. Moreover, we specialize our solution to be robust to traffic classification errors. Extensive numerical experiments characterize the performance of our approach in vitro. Finally, results of the implementation in a real OpenFlow controller demonstrate the viability of SOFIA as a solution in production environments.

The surge of mobile data traffic forces network operators to cope with capacity shortage. The deployment of small cells in 5G networks shall increase radio access capacity. Mobile edge computing technologies can be used to manage dedicated cache memory at the edge of mobile networks. As a result, data traffic can be confined within the radio access network thus reducing latency, round-trip time and backhaul congestion. Such technique can be used to offer content providers premium connectivity services to enhance the quality of experience of their customers on the move. In this context, cache memory in the mobile edge network becomes a shared resource. We study a competitive caching scheme where contents are stored at a given price set by the mobile network operator. We first formulate a resource allocation problem for a tagged content provider seeking to minimize the expected missed cache rate. The optimal caching policy is derived accounting for popularity of contents, spatial distribution of small cells, and caching strategies of competing content providers. Next, we study a game among content providers in the form of a generalized non-smooth Kelly mechanism with bounded strategy sets and heterogeneous players. Existence and uniqueness of the Nash equilibrium are proved. Finally, numerical results validate and characterize the performance of the system.

The endeavor to develop 5G technology aims to support the recent outstanding mobile data traffic growth. In this regard, mobile network providers will be able to leverage on cloud edge-caching to offer services with enhanced quality of experience on the move. By this technology, dedicated cache space of mobile networks can be provisioned to OTT content providers, e.g., over metropolitan areas covered the network of a mobile network provider. In this work we address the problem of fair pricing such caching service, with storage the actual shared resource for caching. We study a scheme in which contents are dynamically stored in the edge memory. The mobile network provider offers a price λ for storing contents on the shared cache, thus engendering competition for cache memory sharing among content providers. We model such competition among OTT content providers using the economic notion of Kelly mechanism. Hence, we have studied the Stackelberg equilibrium, i.e., the optimal price configuration for the network provider. Numerical results describe the structure of the Nash equilibrium and the optimal prices resulting from the network provider optimal strategy.