Network Load Variations Research Articles

Improved Grant-Free Access for URLLC via Multi-Tier-Driven Computing: Network-Load Learning, Prediction, and Resource Allocation

Grant-Free (GF) access has been recognized as a promising candidate for Ultra-Reliable and Low-Latency Communications (URLLC). However, even with GF access, URLLC still may not effectively gain high reliability and millimeter-level latency, simultaneously. This is because the network load is typically time-varying and not known to the base station (BS), and thus, the resource allocated for GF access cannot well adapt to variations of the network load, resulting in low resource utilization efficiency under light network load and leading to severe collisions under heavy network load. To tackle this problem, we propose a multi-tier-driven computing framework and the associated algorithms for URLLC to support users with different QoS requirements. Especially, we concentrate on <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$K $ </tex-math></inline-formula> - repetition GF access in light of its simplicity and well-balanced performance for practical systems. In particular, our framework consists of three tiers of computation, namely network-load learning, network-load prediction, and adaptive resource allocation. In the first tier, the BS can learn the network-load information from the states (success, collision, and idle) of random-access resources in terms of resource blocks (RB) and time slots. In the second tier, the network-load variation is effectively predicted based on estimation results from the first tier. Finally, in the third tier, by deriving and weighing the failure probabilities of different groups of users, their QoS requirements, and the predicted network loads, the BS is able to dynamically allocate sufficient resources accommodating the varying network loads. Simulation results show that our proposed approach can estimate the network load more accurately compared with the baseline schemes. Moreover, with the assistance of network-load prediction, our adaptive resource allocation offers an effective way to enhance the QoS for different URLLC services, simultaneously.

On the Robustness of Controlled Deep Reinforcement Learning for Slice Placement

The evaluation of the impact of using Machine Learning in the management of softwarized networks is considered in multiple research works. In this paper, we propose to evaluate the robustness of online learning for optimal network slice placement. A major assumption in this study is to consider that slice request arrivals are non-stationary. We precisely simulate unpredictable network load variations and compare two Deep Reinforcement Learning (DRL) algorithms: a pure DRL-based algorithm and a heuristically controlled DRL as a hybrid DRL-heuristic algorithm, in order to assess the impact of these unpredictable changes of traffic load on the algorithms performance. We conduct extensive simulations of a large-scale operator infrastructure. The evaluation results show that the proposed hybrid DRL-heuristic approach is more robust and reliable than pure DRL in real network scenarios.

Enhanced network sensitive access control scheme for LTE–LAA/WiFi coexistence: Modeling and performance analysis

Implementation of 5G and beyond networks is looking to expand the operation of licensed systems into unlicensed frequency bands through technologies such as license assisted access (LAA) and New Radio Unlicensed (NR-U). LAA aggregates licensed and unlicensed bands for long term evolution (LTE) implementation to address the ever-increasing demand for cellular data traffic and shortage of licensed spectrum. Spectrum-efficient resource-sharing schemes are, however, critical for the harmonious coexistence of LTE–LAA with incumbent systems on the unlicensed band, especially WiFi, while opportunistically improving LTE throughput and enhancing spectrum utilization of the unlicensed band. In this paper, we present a listen before talk (LBT) based clear channel assessment (CCA) mechanism for LAA eNodeBs (eNBs) to improve the coexistence performance of LTE–LAA with WiFi. Particularly, we propose an adaptive exponential backoff scheme for LTE eNB that dynamically updates the contention window (CW) size and transmission opportunity (TXOP) parameters according to network load variations. A three-dimensional discrete Markov model is developed to describe the LBT procedure of LAA eNB, and a performance model is further derived to evaluate steady-state channel access, transmission, and failure probabilities. The proposed scheme’s performance is evaluated in terms of successful transmission probability, throughput, and delay according to the 3GPP and WiFi guidelines. The results are compared with the traditional schemes proposed in literature considering fixed and adaptive CW size for LAA eNB.

Generalized Proportional Allocation Policies for Robust Control of Dynamical Flow Networks

We study a robust control problem for dynamical flow networks. In the considered framework, traffic flows along the links of a transportation network—modeled as a capacited multigraph—and queues up at the nodes, whereby control policies determine which incoming queues are to be allocated service simultaneously, within some predetermined scheduling constraints. We first prove fundamental performance limitations on the system performance by showing that for a dynamical flow network to be stabilizable by some control policy, it is necessary that the exogenous inflows belong to a certain stability region, that is determined by the network topology, the link capacities, and the scheduling constraints. Then, we introduce a family of distributed controls, referred to as generalized proportional allocation (GPA) policies, and prove that they stabilize a dynamical transportation network whenever the exogenous inflows belong to such stability region. The proposed GPA control policies are decentralized and fully scalable as they rely on local feedback information only. Differently from previously studied maximally stabilizing control strategies, the GPA control policies do not require any global information about the network topology, the exogenous inflows, or the routing, which makes them robust to unpredicted network load variations and changes in the link capacities or the routing decisions. Moreover, the proposed GPA control policies also take into account the overhead time while switching between services. Our theoretical results find one application in the control of urban traffic networks with signalized intersections, where vehicles have to queue up at junctions and the traffic signal controls determine the green light allocation to the different incoming lanes.

Load-Dependent Cascading Failures in Finite-Size Erdös-Rényi Random Networks

Large-scale cascading failures can be triggered by very few initial failures, leading to severe damages in complex networks. This paper studies load-dependent cascading failures in random networks consisting of a large but finite number of components. Under a random single-node attack, a framework is developed to quantify the damage at each stage of a cascade. Estimations and analyses for the fraction of failed nodes are presented to evaluate the time-dependent system damage due to the attack. The results provide guidelines for choosing the load margin to avoid a cascade of failures. Furthermore, the analysis reveals a phase transition behavior in the extent of the damage as the load margin grows, i.e., the fraction of the damaged components drops from near one to near zero over a slight change in the load margin. The critical value of the load margin and the short interval over which such an abrupt change occurs are derived to characterize the network reaction to small network load variations. Our findings provide design principles for enhancing the network resiliency in load-dependent complex networks with practical sizes.

License assisted access-WiFi coexistence with TXOP backoff for LTE in unlicensed band

License assisted access (LAA) was proposed as a solution to the challenge of satisfying high throughput requirement in cellular network with limited licensed band. LAA uses long term evolution (LTE) carrier aggregation technique to combine licensed and unlicensed bands, specifically 5 GHz WiFi band, in order to achieve high data rates. However, LAA has several critical concerns for services provision on 5 GHz unlicensed band. In this article, an overview of the techniques concerning the coexistence of LTE and WiFi on the same unlicensed band is presented. We first present several general approaches including power control of LTE evolved Node B (eNB), carrier selection, LTE duty cycling, self clear to send (CTS) and request to send/CTS (RTS/ CTS), and listen before talk (LBT). Then, we focus on LBT techniques for LAA-WiFi coexistence and discuss how LBT based coexistence solutions can dynamically respond to the variations in network load and ensure fair coexistence in a distributed manner. Based on the LBT mechanism, we also present a fairness-aware LAA-WiFi coexisting scheme with TXOP (transmitted in a single transmission opportunity) backoff for LAA eNB. Finally, the performances of various LBT solutions for LAA-WiFi coexistence are evaluated using Markov chain analytical model.

Dynamic eICIC — A Proactive Strategy for Improving Spectral Efficiencies of Heterogeneous LTE Cellular Networks by Leveraging User Mobility and Traffic Dynamics

Enhanced Inter-cell Interference Co-ordination (eICIC) is a time-domain multiplexing technique for improving the performance of legacy co-channel Heterogeneous Networks (HetNets). eICIC offers several benefits, including a more equitable distribution of user traffic across the macro and embedded pico cells (i.e., load balancing), in turn leading to better pico cell utilizations and better edge user throughputs. eICIC uses two mechanisms to achieve load balancing: (i) a cell selection bias to increase the pico coverage area, enabling the pico to attract more users, and (ii) macro muting ("Almost Blank Sub-frame", ABS) to improve SINRs of range extended pico users. Network load in the cellular system varies continuously as a result of user mobility and traffic dynamics (the varying amount of data pending for a user - arising from packet arrivals and eventual departures). Previous work did not consider dynamically changing network conditions in HetNets. Hence, they studied the benefits of eICIC over legacy HetNets arising from static optimization of eICIC parameters. Our approach is dynamic, and further improves performance by keeping the HetNet continuously optimized for improved user experience, by adapting the ABS in response to dynamic variations in network load. We present several analytical formulations, which allow for simple optimizations and an intuitive understanding of the desired system response to load variation. Exhaustive system simulations illustrate the performance gains along different dimensions, such as spectral efficiency, fairness, mean file transfer times, number of file transmissions, number of file drops, and mean queue lengths. Furthermore, we study the impact of different mobility scenarios, traffic mixes, and adaptation rate. We also provide a brief discussion of practical considerations in implementing Dynamic eICIC.

A Game Incentive Scheme Integrating Punishment in Vehicular Ad Hoc Networks

In vehicular ad hoc networks, uncooperative behaviors will impact the reliability of comfort applications, as well as drivers’ decisions, and even invoke serious traffic accidents. In this paper, we propose a novel game incentive scheme to stimulate cooperation among vehicle nodes, consider selfish nodes’ expectations to future payoff and their long-term desires for profit, and show analytically the three incentive-compatible conditions under which selfish nodes will be deterred from cheating by the subsequent punishments. We also discuss the impact on selfish nodes’ behavior, which is caused by their willingness for future collaboration, the parameter values of punishment mechanism and the variation of network load. Simulation results show that, the increase of network load and the deterioration of node’s future profit expectation will motivate nodes toward self-interested action, but our scheme can neutralize this tendency by the careful configuration of punishment parameters, and have favorable incentive effect.

Uma Avaliação da Capacidade de Transmissão de Voz em Redes 802.11 no Modo Ad-Hoc

Tills paper analyzes voice transmission capacity on ad hoc networks by performing simulations related to delay, jitter, loss rate, and consecutive losses. We evaluatethe influence of mobility and quality of service provisioning on the number of voice transnlitting sources. Another issue addressed in this paper is the effect of node density in voice transmission. Results show that we can achieve up to 22% ofimprovement in the capacity of voice transmission when the medium access time is reduced by service differentiation on the MAC layer. Moreover, mobility and network load variations can degrade up to 60% the network capacity of voice transmission, mainly on multihop mobile networks.

Servicing and Real-Time Control of Networks With Dynamic Routing

The design of a network for dynamic routing is made using the forecasted network loads. Load uncertainties arising from errors in the forecast and from daily variations in network load give rise to reserve or idle network capacity not immediately needed by current network demands. The reserve capacity can be reduced by the use of more flexible dynamic routing methods, which allow routing flexibility to help control network flow under load uncertainties. We illustrate techniques for changing network routing patterns in planned and demand servicing to counteract the effects of forecast errors. Included in the benefits are a reduction in both reserve capacity, estimated to be about 5 percent of network first cost, and in trunk rearrangements. We also present call-by-call simulation results for real-time routing enhancements to the basic routing algorithms. The real-time routing algorithms use dynamic trunk reservation techniques, and the simulation results illustrate the improvement in network efficiency and performance under normal daily load variations, network overloads, and network failures.