Search

We address the resource allocation challenges in NextG cellular radio access networks (RAN), where heterogeneous user applications demand guarantees on throughput and service regularity. We leverage the Whittle indexability property to decompose the resource allocation problem, enabling the independent computation of relative priorities for each user. By simply allocating resources in decreasing order of these indices, we transform the combinatorial complexity of resource allocation into a linear one. We propose Windex, a lightweight approach for training neural networks to compute Whittle indices, considering constraint violation, channel quality, and system load. Implemented on a real-time RAN intelligent controller (RIC), our approach enables resource allocation decision times of less than 20$\mu$s per user and efficiently allocates resources in each 1ms scheduling time slot. Evaluation across standardized 3GPP service classes demonstrates significant improvements in service guarantees compared to existing schedulers, validated through simulations and emulations with over-the-air channel traces on a 5G testbed., Comment: 15 pages, 8 figures

Media streaming is the dominant application over wireless edge (access) networks. The increasing softwarization of such networks has led to efforts at intelligent control, wherein application-specific actions may be dynamically taken to enhance the user experience. The goal of this work is to develop and demonstrate learning-based policies for optimal decision making to determine which clients to dynamically prioritize in a video streaming setting. We formulate the policy design question as a constrained Markov decision problem (CMDP), and observe that by using a Lagrangian relaxation we can decompose it into single-client problems. Further, the optimal policy takes a threshold form in the video buffer length, which enables us to design an efficient constrained reinforcement learning (CRL) algorithm to learn it. Specifically, we show that a natural policy gradient (NPG) based algorithm that is derived using the structure of our problem converges to the globally optimal policy. We then develop a simulation environment for training, and a real-world intelligent controller attached to a WiFi access point for evaluation. We empirically show that the structured learning approach enables fast learning. Furthermore, such a structured policy can be easily deployed due to low computational complexity, leading to policy execution taking only about 15$\mu$s. Using YouTube streaming experiments in a resource constrained scenario, we demonstrate that the CRL approach can increase quality of experience (QOE) by over 30\%., Comment: 15 pages, 14 figures

Safe reinforcement learning is extremely challenging--not only must the agent explore an unknown environment, it must do so while ensuring no safety constraint violations. We formulate this safe reinforcement learning (RL) problem using the framework of a finite-horizon Constrained Markov Decision Process (CMDP) with an unknown transition probability function, where we model the safety requirements as constraints on the expected cumulative costs that must be satisfied during all episodes of learning. We propose a model-based safe RL algorithm that we call Doubly Optimistic and Pessimistic Exploration (DOPE), and show that it achieves an objective regret $\tilde{O}(|\mathcal{S}|\sqrt{|\mathcal{A}| K})$ without violating the safety constraints during learning, where $|\mathcal{S}|$ is the number of states, $|\mathcal{A}|$ is the number of actions, and $K$ is the number of learning episodes. Our key idea is to combine a reward bonus for exploration (optimism) with a conservative constraint (pessimism), in addition to the standard optimistic model-based exploration. DOPE is not only able to improve the objective regret bound, but also shows a significant empirical performance improvement as compared to earlier optimism-pessimism approaches., Comment: Accepted to NeurIPS 2022

Many physical systems have underlying safety considerations that require that the policy employed ensures the satisfaction of a set of constraints. The analytical formulation usually takes the form of a Constrained Markov Decision Process (CMDP). We focus on the case where the CMDP is unknown, and RL algorithms obtain samples to discover the model and compute an optimal constrained policy. Our goal is to characterize the relationship between safety constraints and the number of samples needed to ensure a desired level of accuracy -- both objective maximization and constraint satisfaction -- in a PAC sense. We explore two classes of RL algorithms, namely, (i) a generative model based approach, wherein samples are taken initially to estimate a model, and (ii) an online approach, wherein the model is updated as samples are obtained. Our main finding is that compared to the best known bounds of the unconstrained regime, the sample complexity of constrained RL algorithms are increased by a factor that is logarithmic in the number of constraints, which suggests that the approach may be easily utilized in real systems.

Crucial performance metrics of a caching algorithm include its ability to quickly and accurately learn a popularity distribution of requests. However, a majority of work on analytical performance analysis focuses on hit probability after an asymptotically large time has elapsed. We consider an online learning viewpoint, and characterize the "regret" in terms of the finite time difference between the hits achieved by a candidate caching algorithm with respect to a genie-aided scheme that places the most popular items in the cache. We first consider the Full Observation regime wherein all requests are seen by the cache. We show that the Least Frequently Used (LFU) algorithm is able to achieve order optimal regret, which is matched by an efficient counting algorithm design that we call LFU-Lite. We then consider the Partial Observation regime wherein only requests for items currently cached are seen by the cache, making it similar to an online learning problem related to the multi-armed bandit problem. We show how approaching this "caching bandit" using traditional approaches yields either high complexity or regret, but a simple algorithm design that exploits the structure of the distribution can ensure order optimal regret. We conclude by illustrating our insights using numerical simulations.

The predominant use of wireless access networks is for media streaming applications, which are only gaining popularity as ever more devices become available for this purpose. However, current access networks treat all packets identically, and lack the agility to determine which clients are most in need of service at a given time. Software reconfigurability of networking devices has seen wide adoption, and this in turn implies that agile control policies can be now instantiated on access networks. The goal of this work is to design, develop and demonstrate QFlow, a learning approach to create a value chain from the application on one side, to algorithms operating over reconfigurable infrastructure on the other, so that applications are able to obtain necessary resources for optimal performance. Using YouTube video streaming as an example, we illustrate how QFlow is able to adaptively provide such resources and attain a high QoE for all clients at a wireless access point., Comment: Submitted to ToN in May, 2020

In this paper, we study the performance of a TCP connection over cognitive radio networks. In these networks, the network may not always be available for transmission. Also, the packets can be lost due to wireless channel impairments. We evaluate the throughput and packet retransmission timeout probability of a secondary TCP connection over an ON/OFF channel. We first assume that the ON and OFF time durations are exponential and later extend it to more general distributions. We then consider multiple TCP connections over the ON/OFF channel. We validate our theoretical models and the approximations made therein via ns2 simulations.