Traffic and Performance Engineering for Heterogeneous Networks

Performance modeling is important for the purpose of developing efficient dimensioning tools for large complicated networks. But it is difficult to achieve in heterogeneous wireless networks, where different networks have different statistical characteristics in service and traffic models. Multiservice loss analysis based on multi-dimensional Markov chain becomes intractable in these networks due to intensive computations required. This paper focuses on performance modeling for heterogeneous wireless networks based on a hierarchical overlay infrastructure. A method based on decomposition of the correlated traffic is used to achieve an approximate performance modeling for multiservice in hierarchical heterogeneous wireless networks with overflow traffic. The accuracy of the approximate performance obtained by our proposed modeling is verified by simulations.

Traffic engineering can be viewed as the mechanisms that are used to reoptimise network performance on traffic variation time scales. The traffic engineering process involves measurements, detection of performance problems, and the adjustment of routes and various quality of service mechanisms in order to restore the desired performance. Traffic engineering models are essential in this feedback loop. In this paper we provide an overview of our recent contributions to analytical models for traffic engineering of elastic traffic in internets, and discuss the role that such models could play in automated or intelligent traffic engineering.

Performance modeling and evaluation of heterogeneous computer networks

Supporting URLLC has become one of the major considerations in the design of 5G systems. In the literature, it has been shown that cooperative relaying is an efficient strategy to improve the reliability of transmissions, support higher rates, and lower latency. However, prior studies have demonstrated the performance advantages of relaying generally under the ideal assumption of communicating arbitrarily reliably at Shannon's channel capacity, which is not an accurate performance indicator for relaying in URLLC networks in which transmission is required to be completed within a strict time span and coding schemes with relatively short blocklengths need to be employed. In this article, we address the performance modeling and optimization of relaying-enabled URLLC networks. We first discuss the accurate performance modeling of relay-enabled 5G networks. In particular, we provide a comprehensive summary of the performance advantage of applying relaying in 5G URLLC transmissions in comparison to the case of direct transmission (without relaying). Both a noise-limited scenario and an interference- limited scenario are discussed. Then we present tools for performance optimization utilizing the knowledge of either perfect or average channel side information. Finally, we summarize the proposed optimization schemes and discuss potential future research directions.

Asynchronous Transfer Mode (ATM) has been chosen as the multiplexing technique for Broadband Integrated Services Digital Networks. Under ATM, all information types (voice, data, image and video) are represented in terms of standardized units called cells. The acceptance of the ATM cell by the computer and communications industry as standard unit of information exchange will have a significant impact on the future development of communications infrastructure and end-user equipment. In this paper, we consider the performance modeling and evaluation of ATM-based networks. The performance issues at the ATM Layer and the ATM Adaptation Layer (AAL) are discussed. These include traffic characterization, multiplexing and switching, quality of service classes, call acceptance, congestion control, and AAL performance. Performance results on multiplexing and switching, and LAN interconnection using an ATM switch, are presented.

Although OSNs are major and growing large scale web applications, there is still a lack of workload models and tools for performance evaluation and testability studies. This fact motivates us to develop a general purpose benchmark for evaluating the main hardware and software resources associated to this kind of applications. To this end, we have developed a flexible workload model based on interactive users that, together with a complete and fully operative framework, permits to monitor system resources to perform fine grain performance and testability studies.1

As a result of its success, Wi-Fi has been promoted as an access technology to provide full internet connectivity in the municipal area. Supporting application QoS by managing resource provisioning is an important challenge for wireless network operators. This paper proposes a traffic engineering model for managing a Voice over Wi-Fi service in a municipal area with stationary and mobile clients. Our proactive algorithm guarantees the connectivity of the mobile clients while maximizing the number of stationary stations. Simulation results show substantial QoS improvements for mobile clients and a near optimal utilization of the capacity of the wireless access network next to the congested hotspots.

Hyper-Dense Heterogeneous (HDH) deployments, which are made of different kinds of Base Stations (BSs), constitute a promising solution to meet the high data rates envisioned for 5G cellular systems. However, the presence of such a large number of BSs, each of which is expected to deliver a large amount of data to the end-users, is pushing the bottleneck of cellular networks from the Radio Access Network (RAN) to the backhaul, which is becoming as critical as the radio infrastructure. Hence, new network topologies and network architectures are under debate, whose objective is to overcome the backhaul cost and capacity crunch. In order to adequately design these emerging network topologies, appropriate abstraction models need to be introduced for system-level analysis. In this context, stochastic geometry has recently emerged as a promising tool for system-level performance evaluation of cellular systems. In this paper, we adopt a Poisson Tree model for analyzing a hierarchical backhaul and the RAN of a HDH network, where the nodes of the tree are traffic concentrators, BSs and users. The proposed model captures the impact of the finite user density (load) on the network performance and the distribution of the Signal-to-Interference-plus-Noise-Ratio (SINR), which is determined by the positions of users and BSs, as well as by the channel conditions and the other-cell interference.

Hybrid routing approaches which either combine the best of traditional IGP and MPLS routing or allow a smooth migration from traditional IGP routing to the newly proposed MPLS standard have scarcely been addressed by the IP community. This paper presents an on-line traffic engineering (TE) model which uses a hybrid routing approach to achieve efficient routing of flows in IP networks. The model assumes a network-design process where the IP flows are classified at the ingress of the network and handled differently into the core using a multiple metric routing mechanism leading to the logical separation of a physical network into two virtual networks: An IGP network carrying low bandwidth demanding (LBD) flows and an MPLS network where bandwidth-guaranteed tunnels are setup to route high bandwidth demanding (HBD) flows. The hybrid routing approach uses a route optimization model where (1) link weight optimization (LWO) is implemented in the IGP network to move the IP flows away from links that provide a high probability to become bottleneck for traffic engineering and (2) the MPLS network is engineered to minimize the interference among competing flows and route the traffic away from heavily loaded links. We show that a cost-based optimization framework can be used by an ISP to design simple and flexible routing approaches where different metrics reflecting the ISP's view of its TE objectives are deployed to improve the use and efficiency of a network. This is achieved through extensions to a mixed cost metric derived from the route optimization model. The IGP+MPLS routing approach is applied to compute paths for the traffic offered to a 15- and 50-node network. Preliminary simulation reveals performance improvements compared to both IGP and MPLS routing in terms of the routing optimality and the network reliability.

Design and performance analysis of an inductive QoS routing algorithm

This SpringerBrief provides an in-depth look at the key issues that affect the performance of heterogeneous networks and presents schemes that can effectively tackle these issues. In particular, this book discusses unbalanced traffic load among the macro and micro Base Stations (BSs) caused by the transmit power disparity, and a load-balancing based mobile association scheme to balance the traffic load among the macro and micro BSs. This book also introduces a fractional frequency reuse (FFR) scheme with proper power control to help reduce interference at the UEs which are most vulnerable to such intra-cell interference. The last section investigates radio resource allocation issues for heterogeneous networks with cooperative relays, and proposes a resource allocation framework that could achieve proportional fairness among the UEs. Numerical results are provided to demonstrate the effectiveness of the proposed solutions in tackling the problem and improving network performance. Resource Management for Heterogeneous Networks in LTE-A Systems is designed for researchers and professionals working in networking and resource management. The content is also valuable for advanced-level students in computer science and electrical engineering.

In this paper, a stochastic traffic assignment model for networks is proposed for the study of discrete dynamic Bayesian algorithms. In this paper, we study a feasible method and theoretical system for implementing traffic engineering in networks based on Bayesian algorithm theory. We study the implementation of traffic assignment engineering in conjunction with the network stochastic model: first, we study the Bayesian algorithm theoretical model of control layer stripping in the network based on the discrete dynamic Bayesian algorithm theory and analyze the resource‐sharing mechanism in different queuing rules; second, we study the extraction and evaluation theory of traffic assignment for the global view obtained by the control layer of the network and establish the Bayesian algorithm analysis model based on the traffic assignment; subsequently, the routing of bandwidth guarantee and delay guarantee in the network is studied based on Bayesian algorithm model and Bayesian algorithm network random traffic allocation theory. In this paper, a Bayesian algorithm estimation model based on Bayesian algorithm theory is constructed based on network random observed traffic assignment as input data. The model assumes that the roadway traffic distribution follows the network random principle, and based on this assumption, the likelihood function of the roadway online traffic under the network random condition is derived; the prior distribution of the roadway traffic is derived based on the maximum entropy principle; the posterior distribution of the roadway traffic is solved by combining the likelihood function and the prior distribution. The corresponding algorithm is designed for the model with roadway traffic as input, and the reliability of the algorithm is verified in the arithmetic example.

Recurring congestion on highways is the primary cause of traffic congestion in the urban traffic system while jams on merging sections are the worst section along the expressway. Due to the presence of heavy congestion in the site of University of Technology, this study aims to discover the traffic characteristics of this site to determine the appropriate solutions. Therefore, this study selects a merging section on Mohammed Al-Qassim expressway to investigate some characteristics of such a section in Iraq. Field data have been collected from this site for 4 h. The data were collected from video cameras installed along the expressway in Baghdad city which is called Mohammed Al-Qassim Expressway. These data present flow of upstream and ramp. The Advanced Interactive Microscopic Simulator for Urban and Non-Urban Networks (AIMSUN) program was also adopted to investigate other characteristics which are difficult to get such as speed and occupancy for the current case. The results showed that the simulated and real data significantly corresponded. In this study, the AIMSUN program received recognition by utilizing Geoffrey E. Havers a method for comparing two sets of traffic volumes that are used in traffic engineering, traffic forecasting, and traffic modeling as it was discovered to be comparable to reality, and the road management was then enhanced by the addition of ramp metering and cycle (green) time control.

As network traffic becomes increasingly dynamic and service demands grow more diverse, achieving accurate traffic visibility and responsive path control has become a central challenge in traffic engineering. This paper proposes a unified SDN architecture tailored for SRv6-enabled networks, which integrates traffic matrix reconstruction, congestion-aware rerouting, and strategy enforcement into a closed-loop control system. The architecture consists of four functional layers: data collection, traffic modeling, path optimization, and policy deployment. A lightweight inference method based on SRv6 Prefix-SID counters is introduced to enable high-accuracy traffic matrix reconstruction with minimal overhead. To address congestion events, a load minimizing rerouting algorithm is designed using linear programming techniques. Additionally, SRv6 forwarding policies are automatically generated from YANG models and delivered to edge routers, enabling rapid and version-controlled policy deployment. Experimental results show that the proposed method reduces the reconstruction error by up to 80 % compared with baseline approaches, decreases maximum link utilization by 19.8 % after rerouting, and significantly reduces the number of congested links, thereby improving system accuracy, efficiency, and robustness.

Cell Loss Ratio estimation is a crucial technology in call admission control and traffic engineering in ATM networks. Cell arrival process and analytical approach had been investigated in recent years. However, an effective and practicable CLR estimator is still a challenge. Based on the analysis of cell loss problem, an improved simple traffic model and a new algorithm are presented in this paper, to estimate CLR of ATM network multiplexed with heterogenous traffic classes services rapidly. Traffic model is constructed with standard parameters, so it is easy to use in practical situations. The new algorithm runs quickly enough to respond the call real-time. Simulation results show that accuracy, complexity and robustness of algorithm are ideal to be utilized in real network.