Router Capacity Research Articles

Impact of Segmentation and Popularity-based Cache Replacement Policies on Named Data Networking

The data distribution mechanism of internet protocol (IP) technology is inefficient because it necessitates the user to await a response from the server. Named data networking (NDN) is a cutting-edge technology being assessed for enhancing IP networks, primarily because it incorporates a data packet caching technique on every router. However, the effectiveness of this approach is highly dependent on the router's content capacity, thus requiring the use data replacement mechanism when the router capacity is full. The least recently used (LRU) method is employed for cache replacement policy; yet, it is considered ineffective as it neglects the content's popularity. The LRU algorithm replaces the infrequently requested data, leading to inefficient caching of popular data when multiple users constantly request it. To address this problem, we propose a segmented LRU (SLRU) replacement strategy that considers content popularity. The SLRU will evaluate both popular content and content that has previously been popular in two segment categories, namely the probationary and protected segments. Icarus simulator was used to evaluate multiple comprehensive scenarios. Our experimental results show that the SLRU obtains a better cache hit ratio (CHR) and able to minimize latency and link load compared to existing cache replacement policies such as First In, First Out (FIFO), LRU, and Climb.

Impact of SDN on Optical Router Bypass

Internet traffic is growing rapidly, driven by new applications such as the Internet of things (IoT). To overcome router scaling limitations, router bypass has been introduced. Traditional bypass is based on provisioning fixed-size bypass channels, which was shown to underutilize link capacity. In software defined networking (SDN), router bypass can be requested as a service by the application layer. This study explores the opportunity to bypass routers efficiently without the link underutilization of the traditional bypass.We combine SDN and dynamic layer provisioning to implement dynamic router bypassing and show how this leads to improved efficiency. A simulation built to compare the traditional bypass with the SDN-based bypass showed that SDN-based bypassing enhanced efficiency by 33%. When bypass traffic is increased, we show that the output capacity of a node increases up to 56%, limited by optical capacity rather than router capacity.

RouteBricks

We revisit the problem of scaling software routers, motivated by recent advances in server technology that enable highspeed parallel processing a feature router workloads appear ideally suited to exploit. We propose a software router architecture that parallelizes router functionality both across multiple servers and across multiple cores within a single server. By carefully exploiting parallelism at every opportunity, we demonstrate a 40Gbps parallel router prototype; this router capacity can be linearly scaled through the use of additional servers. Our prototype router is fully programmable using the familiar Click/Linux environment and is built entirely from off-the-shelf, general-purpose server hardware. We also describe some of the lessons learned while supporting field deployments of Routebricks-based software routers.

Distributed bandwidth allocation based on alternating evolution algorithm

In a network, end nodes have to compete for bandwidth through some distributed congestion control algorithms. It is a great challenge to ensure the efficiency and fairness of the distributed control algorithms. TCP congestion control algorithms do not perform well in terms of their efficiency and fairness in high speed networks. In this paper, we propose a novel asymptotic evolution algorithm based on the Logistic Model to allocate limited bandwidth resource. The algorithm introduces an explicit bandwidth pre-allocation factor. The factor is carried by the packet and is computed in routers based on the information of the router capacity, the aggregate load, and the instantaneous queue length; therefore the algorithm does not require the routers to keep the per-flow state. According to this pre-allocation bandwidth factor, the senders asymptotically adjust their sending rate and the bandwidth factor changes asymptotically along with the variation of the aggregate load and the queue length in the routers; therefore the sending rate and the pre-allocation bandwidth factor form alternating evolution and eventually reach a steady state. Theoretical analysis and simulation experiments were conducted to compare our algorithm with related ones. The results show that our algorithm not only provides fast convergence to efficiency and fairness, but also keeps a strong robustness against crossing traffic.

Scaling Virtual Private Networks

Many organizations purchase connectionless Virtual Private Network (VPN) services from a telecommunications service provider, who deploys a network of routers, interconnected by high bandwidth trunks, which serve thousands of customers. The nodes in a service provider network are Provider Edge routers (PEs), which store forwarding tables which dictate on which outgoing link to send an incoming packet addressed to a particular destination. In practice, router capacity may be limited by memory, bandwidth, or processing capacity for BGP (Border Gateway Protocol) sessions used to advertise addresses. We describe three techniques that help a service provider deploy a large-scale VPN service. The first technique shows how a service provider can deploy sets of routers, called “route reflector planes,” to advertise addresses when the number of addresses in the service provider network grows. The second technique, applicable when the number of BGP sessions is the limiting resource, is a method to enhance the network reliability at no additional cost, by having each customers forwarding table stored on only a small number of PEs. The third technique is a way to estimate the number of VPNs that can be handled by a PE, when memory is the limiting resource.

Scheduling and Admission Control in Power-Constrained OFDM Wireless Mesh Routers: Analysis and Optimization

We investigate the packet-level and the connection-level performances in a power-constrained (e.g., solar-powered) wireless mesh router with packet scheduling and admission control. The system model under consideration is compatible with the IEEE 802.16a standard with orthogonal frequency division multiple access (OFDMA) air interface. In the medium access control (MAC) layer, transmission frames are grouped into super-frames and the allocation of subchannel and number of frames to each connection (i.e., scheduling) is performed to satisfy the traffic requirement of each connection and also the power supply constraint at a mesh router/802.16a base station (BS). A queueing analytical model based on discrete-time Markov chain (DTMC) is used to analyze the packet-level performances. Based on this scheduling, router capacity in terms of the maximum number of ongoing connections is obtained. Subsequently, a threshold-based admission control method is proposed for both relay and local connections in a router so that connection-level performances are satisfied. To this end, we optimize the admission control over multiple time periods in which the amount of supplied power and the traffic load at a mesh router are time-dependent.

Design of stochastic hitless‐prediction router by using the first exceed level theory

AbstractThis paper deals with an enhanced hitless‐prediction router system that has the hitless‐restart capability with forecasting. Hitless‐restart means that the router can stay on the forwarding path and the network topology remains stable. But the major difficulty of the current hitless‐restart is that the router is always active to take the action, such as non‐stop forwarding (upgrade, maintenance and capacity expansion may be included as third party activities). Stochastic hitless‐prediction model gives the decision making factors that manage a router system more efficiently. An analogue of the first exceed level theory is applied for the restriction of the number of buffer size that is the router capacity. Analytically, tractable results are obtained by using a first exceed level process that enables us to determine the decision making factors such as recycle periods of the hitless‐prediction point to prevent a router shutdown. Copyright © 2005 John Wiley & Sons, Ltd.

Using Hardware to Configure a Load-Balanced Switch

Efficient router architectures should have predictable throughput and scalable capacity, as well as internal optical technology (such as optical switches and wavelength division multiplexing) that can increase capacity by reducing power consumption. The load-balanced switch is a promising way to scale router capacity. In this 100-terabit-per-second router, an optical switch spreads traffic evenly among linecards. When the network operator adds or removes linecards, reconfiguring the switch can be time consuming, but a polynomial-time algorithm drastically reduces the required memory-intensive operations, yielding a switch-reconfiguration time below 50 ms.

A high‐capacity switching method for IP datagrams with multistage networks

AbstractThe rapidly spreading use of the Internet and the general use of video communication through the Internet have caused a rapid increase in Internet traffic. It is highly probable that the Internet will be used as the network for future multimedia communications. In that case, the capacity of IP routers for the Internet must be enlarged. At present, an IP router capacity of several gigabits per second can be realized, but ATM technology must be applied if a still larger capacity is to be achieved. The IP traffic is of variable length, and the packets are long, which can be handled by ATM. If a high‐capacity switch that can handle packets of variable length is created, performance will be improved compared to high‐capacity switches based on ATM. The purpose of this paper is to achieve a capacity of 1 Tbps order. Several IP datagram switching methods are proposed and evaluated. © 2002 Wiley Periodicals, Inc. Electron Comm Jpn Pt 1, 85(8): 57–65, 2002; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ecja.1118