We present system-level evaluation results to characterize the performance of eXtended Reality (XR) applications over a 5G NR system. A split-rendering framework is assumed, where the XR device uses 5G to offload computation associated with rendering and encoding of video frames to an edge server. The XR traffic model includes rendered video frames on the downlink and user pose and control updates on the uplink. We present results for both sub-6 GHz (FR1) and millimeter-wave (FR2) bands. Additionally, we also discuss potential enhancements to further improve user experience for XR applications over 5G.

The fifth generation (5G) new radio (NR) of mobile communications is designed to support two major class of services with vastly heterogeneous requirements: ultra-reliable low-latency communication (URLLC) and enhanced mobile broadband (eMBB). To enable the coexistence of URLLC and eMBB in the same radio spectrum in a cost-effective manner while guaranteeing the latency and reliability performance of URLLC, 5G NR has developed an innovative preemption and superposition framework which allows a dynamic multiplexing of the two services in an interleaved manner. In this tutorial paper, we review the key technological innovations in this framework in NR Release 15 and 16, namely, preemption indication and enhanced power control, and explain the design principles underlying these technologies. System-level and link-level simulations are provided to illustrate the performance benefits of these designs.

Tactile internet is defined as a network that offers a response to a physical process or an object in perceived real time. The development of new radio (NR) and long-term evolution (LTE) technologies with tailored high-reliability and low-latency design is expected to reliably transmit data in milliseconds, leading to the promising realization of tactile internet with applications in areas such as factory automation, education, gaming, and healthcare. In this paper, from a physical layer and medium-access control (PHY/MAC) perspective, we discuss the systems design of ultrareliable and low-latency communications (URLLC) in NR and LTE technologies, both belonging to the fifth-generation (5G) wireless technologies. Motivated by the safety and privacy requirements of tactile internet, we also outline the 5G security landscape, major categories of attackers, and potential countermeasures. Finally, we provide a case study of factory automation as an example of the proposed 5G-based tactile internet.

The backpressure routing policy is known to be a throughput optimal policy that supports any feasible traffic demand, but may have poor delay performance when packets traverse loops in the network. In this paper, we study loop-free backpressure routing policies that forward packets along directed acyclic graphs (DAGs) to avoid the looping problem. These policies use link reversal algorithms to improve the DAGs in order to support any achievable traffic demand. For a network with a single commodity, we show that a DAG that supports a given traffic demand can be found after a finite number of iterations of the link-reversal process. We use this to develop a joint link-reversal and backpressure routing policy, called the loop free backpressure (LFBP) algorithm. This algorithm forwards packets on the DAG, while the DAG is dynamically updated based on the growth of the queue backlogs. We show by simulations that such a DAG-based policy improves the delay over the classical backpressure routing policy. We also propose a multicommodity version of the LFBP algorithm and via simulation show that its delay performance is better than that of backpressure.

We study the problem of efficiently disseminating packets in multi-hop wireless networks. At each time slot, the network controller activates a set of non-interfering links and forward selected copies of packets on each activated link. The maximum rate of commonly received packets is referred to as the broadcast capacity of the network. Existing policies achieve the broadcast capacity by balancing traffic over a set of spanning trees, which are difficult to maintain in a large and time-varying wireless network. In this paper, we propose a new dynamic algorithm that achieves the broadcast capacity when the underlying network topology is a directed acyclic graph (DAG). This algorithm is decentralized, utilizes local information only, and does not require the use of spanning trees. The principal methodological challenge inherent in this problem is the absence of work-conservation principle due to the duplication of packets, which renders usual queuing modeling inapplicable. We overcome this difficulty by studying relative packet deficits and imposing in-order delivery constraints to every node in the network. We show that in-order delivery is throughput-optimal in DAGs and can be exploited to simplify the design and analysis of optimal algorithms. Our capacity characterization also leads to a polynomial time algorithm for computing the broadcast capacity of any wireless DAG under the primary interference constraints. In addition, we propose a multiclass extension of our algorithm, which can be effectively used for broadcasting in any network with arbitrary topology. Simulation results show that the our algorithm has a superior delay performance as compared with the traditional tree-based approaches.

We present a novel control scheme that dynamically optimizes multirate multicast. By computing the differential backlog at every node, our scheme adaptively allocates transmission rates per session/user pair in order to maximize throughput. An important feature of the proposed scheme is that it does not require source cooperation or centralized calculations. This methodology leads to efficient and distributed algorithms that scale gracefully and can be embraced by low-cost wireless devices. Additionally, it is shown that maximization of sum utility is possible by the addition of a virtual queue at each destination node of the multicast groups. The virtual queue captures the desire of the individual user and helps in making the correct resource allocation to optimize total utility. Under the operation of the proposed schemes backlog sizes are deterministically bounded, which provides delay guarantees on delivered packets. To illustrate its practicality, we present a prototype implementation in the NITOS wireless testbed. The experimental results verify that the proposed schemes achieve maximum performance while maintaining low complexity.

We study the problem of broadcasting packets in wireless networks. At each time slot, a network controller activates non-interfering links and forwards packets to all nodes at a common rate; the maximum rate is referred to as the broadcast capacity of the wireless network. Existing policies achieve the broadcast capacity by balancing traffic over a set of spanning trees, which are difficult to maintain in a large and time-varying wireless network. We propose a new dynamic algorithm that achieves the broadcast capacity when the underlying network topology is a directed acyclic graph (DAG). This algorithm utilizes local queue-length information, does not use any global topological structures such as spanning trees, and uses the idea of in-order packet delivery to all network nodes. Although the in-order packet delivery constraint leads to degraded throughput in cyclic graphs, we show that it is throughput optimal in DAGs and can be exploited to simplify the design and analysis of optimal algorithms. Our simulation results show that the proposed algorithm has superior delay performance as compared to tree-based approaches.

The backpressure routing policy is known to be a throughput optimal policy that supports any feasible traffic demand in data networks, but may have poor delay performance when packets traverse loops in the network. In this paper, we study loop-free backpressure routing policies that forward packets along directed acyclic graphs (DAGs) to avoid the looping problem. These policies use link reversal algorithms to improve the DAGs in order to support any achievable traffic demand.

We consider utility maximization in networks where the sources do not employ flow control and may consequently overload the network. In the absence of flow control at the sources, some packets will inevitably have to be dropped when the network is in overload. To that end, we first develop a distributed, threshold-based packet-dropping policy that maximizes the weighted sum throughput. Next, we consider utility maximization and develop a receiver-based flow control scheme that, when combined with threshold-based packet dropping, achieves the optimal utility. The flow control scheme creates virtual queues at the receivers as a push-back mechanism to optimize the amount of data delivered to the destinations via back-pressure routing. A new feature of our scheme is that a utility function can be assigned to a collection of flows, generalizing the traditional approach of optimizing per-flow utilities. Our control policies use finite-buffer queues and are independent of arrival statistics. Their near-optimal performance is proved and further supported by simulation results.

We consider the problem of transmitting multicast flows with hard deadlines over unreliable wireless channels. Every user in the network subscribes to several multicast flows, and requires a minimum throughput for each subscribed flow to meet the QoS constraints. The network controller schedules the transmissions of multicast traffic based on the instant feedback from the users. We characterize the multicast throughput region by analyzing its boundary points, each of which is the solution to a finite-horizon dynamic programming problem over an exponentially large state space. Using backward induction and interchange arguments, we show that the dynamic programming problems are solved by greedy policies that maximize the immediate weighted sum throughput in every slot. Furthermore, we develop a dynamic throughput-optimal policy that achieves any feasible throughput vector by tracking the running performance received by the users.

Multirate multicast improves user quality but complicates network optimization. This paper introduces a novel control scheme to dynamically optimize multirate multicast. We present MMT, an adaptive policy which combines differential backlog scheduling and intelligent packet dropping, both based on local information. MMT is shown to maximize network throughput by adapting to changing conditions such as channel quality, network congestion, and device capabilities. Then, we study the problem of per-receiver network utility maximization. To maximize sum utility we propose the MMU policy, an extension of MMT with receiver-end flow control. Under the operation of both policies backlog sizes are deterministically bounded, which provides delay guarantees on delivered packets. An important feature of the proposed scheme is that it does not require source cooperation or centralized calculations. To illustrate its practicality, we present a prototype implementation in the NITOS wireless testbed. Experimental results verify the optimality of the scheme and its low complexity.

We consider the problem of optimal load balancing in a server farm under overload conditions. A convex penalty minimization problem is studied to optimize queue overflow rates at the servers. We introduce a new class of α-fair penalty functions, and show that the cases of α = 0, 1, ∞ correspond to minimum sum penalty, penalty proportional fairness, and min-max fairness, respectively. These functions are useful to maximize the time to first buffer overflow and minimize the recovery time from temporary overload. In addition, we show that any policy that solves an overload minimization problem with strictly increasing penalty functions must be throughput optimal. A dynamic control policy is developed to solve the overload minimization problem in a stochastic setting. This policy generalizes the well-known join-the-shortest-queue (JSQ) policy and uses intelligent job tagging to optimize queue overflow rates without the knowledge of traffic arrival rates.

We consider utility maximization in networks where the sources do not employ flow control and may consequently overload the network. In the absence of flow control at the sources, some packets will inevitably have to be dropped when the network is in overload. To that end, we first develop a distributed, threshold-based packet dropping policy that maximizes the weighted sum throughput. Next, we consider utility maximization and develop a receiver-based flow control scheme that, when combined with threshold-based packet dropping, achieves the optimal utility. The flow control scheme creates virtual queues at the receivers as a push-back mechanism to optimize the amount of data delivered to the destinations via back-pressure routing. A novel feature of our scheme is that a utility function can be assigned to a collection of flows, generalizing the traditional approach of optimizing per-flow utilities. Our control policies use finite-buffer queues and are independent of arrival statistics. Their near-optimal performance is proved and further supported by simulation results.

We study two convex optimization problems in a multi-class M/G/1 queue with adjustable service rates: minimizing convex functions of the average delay vector, and minimizing average service cost, both subject to per-class delay constraints. Using virtual queue techniques, we solve the two problems with variants of dynamic cμ rules. These algorithms adaptively choose a strict priority policy, in response to past observed delays in all job classes, in every busy period. Our policies require limited or no statistics of the queue. Their optimal performance is proved by Lyapunov drift analysis and validated through simulations.

This paper considers maximizing throughput utility in a multi-user network with partially observable Markov ON/OFF channels. Instantaneous channel states are never known, and all control decisions are based on information provided by ACK/NACK feedback from past transmissions. This system can be viewed as a restless multi-armed bandit problem with a concave objective function of the time average reward vector. Such problems are generally intractable. However, we provide an approximate solution by optimizing the concave objective over a non-trivial inner bound on the network performance region, where the inner bound is constructed by randomizing well-designed stationary policies. Using a new frame-based Lyapunov drift argument, we design a policy of admission control and channel selection that stabilizes the network with throughput utility that can be made arbitrarily close to the optimal in the inner performance region. Our problem has applications in limited channel probing in wireless networks, dynamic spectrum access in cognitive radio networks, and target tracking of unmanned aerial vehicles. Our analysis generalizes the MaxWeight-type scheduling policies in stochastic network optimization theory from time-slotted systems to frame-based systems that have policy-dependent frame sizes.

We consider a wireless base station serving L users through L time-varying channels. It is well known that opportunistic scheduling algorithms with full channel state information (CSI) can stabilize the system with any data rates within the capacity region. However, such opportunistic scheduling algorithms may not be energy efficient when the cost of channel acquisition is high and traffic rates are low. In particular, under the low traffic rate regime, it may be sufficient and more energy efficient to transmit data with no CSI, i.e., to transmit data blindly, since no power for channel acquisition is consumed. In general, we show strategies that probe channels in every slot or never probe channels in any slot are not necessarily optimal, and we must consider mixed strategies. We derive a unified scheduling algorithm that dynamically chooses to transmit data with full or no CSI based on queue backlog and channel statistics. Our methodology is general and can be naturally extended to include timing overhead due to channel acquisition, and to treat systems that allow any subset of channels to be measured. Through Lyapunov analysis, we show that the unified algorithm is throughput-optimal and stabilizes the downlink with optimal power consumption, balancing well between channel-aware and channel-blind transmission modes.

We consider optimal control for general networks with both wireless and wireline components and time varying channels. A dynamic strategy is developed to support all traffic whenever possible, and to make optimally fair decisions about which data to serve when inputs exceed network capacity. The strategy is decoupled into separate algorithms for flow control, routing, and resource allocation, and allows each user to make decisions independent of the actions of others. The combined strategy is shown to yield data rates that are arbitrarily close to the optimal operating point achieved when all network controllers are coordinated and have perfect knowledge of future events. The cost of approaching this fair operating point is an end-to-end delay increase for data that is served by the network. Analysis is performed at the packet level and considers the full effects of queueing.

Abstract — We investigate the combination of distributed ge-ographic routing with transmission power control for energy efficient delivery of information in multihop wireless networks. Using realistic models for wireless channel fading as well as radio modulation and encoding, we first show that the optimal power control strategy over a given link should set the transmission power to achieve a special signal-to-noise ratio (SNR) constant that can be computed using an elegant characteristic equation. Counter-intuitively, for typical radios, this corresponds to an optimal operating point of SNR that lies in the transitional region (where packet error rates are non-negligible). We then propose a local power efficiency metric for distributed routing such that at each step the transmitter picks as the next hop the neighbor for which this metric is maximized. Through extensive simulations, we compare the performance of the proposed algorithm and globally optimal routing algorithms. We show that in randomly deployed 2-D networks, the combination of this local metric for routing with optimal power control has close performances, in terms of average power consumption under different node density settings and physical transmission power limits, to the best strategy using global network link state information. In particular, when electronic power is relatively low, the proposed algorithm can provide up to six times reduction in power usage compared to channel-unaware routing algorithms. I.