Per-packet Overhead Research Articles

Universal and Accurate Sketch for Estimating Heavy Hitters and Moments in Data Streams

In computer networks, traffic measurement is a module in a network probe to measure flow-level statistics from an IP packet stream, which are the basis for network performance monitoring and malicious activity detection. This module extracts the flow IDs from incoming IP packets, classifies packets into flows, and counts the number of packets (or bytes) for each flow. It is a great challenge to measure the per-flow statistics for a high-speed network device, using only the size-limited SRAM on its line cards. Therefore, many algorithms using sublinear memory have been proposed, such as CountMin and CountSketch. However, most of previous algorithms are designed for specific measurement tasks. To obtain multiple types of statistics, people have to deploy multiple sketches, which demands more resources of a network device. It is useful to design a universal sketch that can track not only the top- <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> largest individual flows (called heavy hitters) but also the overall traffic distribution statistics (called moments). Prior work named UnivMon successfully tackled this ambitious quest. However, it incurs large and variable per-packet processing overhead, which may result in a significant throughput bottleneck in high-rate packet stream, given that each packet requires 33 hashes and 32 memory accesses on average and many times of that in the worst case. To address this performance issue, we fundamentally redesign the solution architecture from hierarchical sampling to new progressive sampling and from CountSketch to new GenericCM, which ensure that per-packet overhead is a small constant (5 hashes and 8 memory accesses in the worst case), making it more suitable for online operations, especially for hardware pipeline implementation. This new design also makes effort to reduce memory footprint or equivalently improve measurement accuracy under the same memory. Our experiments show that our solution reduces measurement error by roughly 98.1% for second-order moment and by 91.5% for entropy, when given the same 0.2MB memory as UnivMon.

Empirical Analysis of Extended QX-MAC for IOT-Based WSNS

The Internet of Things (IoT) connects our world in more ways than we imagine. Wireless sensor network (WSN) technology is at the core of implementing IoT architectures. Although WSN applications give us enormous opportunities, their deployment is challenging because of the energy constraint in sensor nodes. The primary design objective of WSNs is therefore to maximize energy efficiency. Enhancing network quality of service (QoS), such as latency, is another crucial factor, particularly for different delay-sensitive applications. Medium access control (MAC) protocols are of paramount importance to achieve these targets. Over the years, several duty-cycled MAC protocols were proposed. Among them, the strobed preamble approach introduced in X-MAC has gained much interest in IoT field because of its several theoretical advantages. However, X-MAC is highly efficient only under light traffic. Under heavy traffic, X-MAC incurs high per-packet overhead and extra delay. In addition, X-MAC has several design flaws that can significantly degrade network performance. In this paper, we point out some specific malfunctions in the original X-MAC design and propose alternatives to reduce their impact. We present an energy-efficient, traffic-adaptive MAC protocol called QX-MAC that addresses the foreseen shortcomings in X-MAC. QX-MAC integrates Q-learning and the more bit scheme to enable the nodes to adapt the active period and duty cycle in accordance with incoming traffic. Finally, the performance of QX-MAC is thoroughly analyzed compared with other reference protocols to validate its efficacy. Our QX-MAC simulation results demonstrate substantial improvements in overall network performance in terms of energy consumption, packet loss, delay, or throughput.

Loop Free Convergence Achieved by Integrating LOLS with FCFR and Gives Lower Convergence Time than OSPF

The short-lived failures are reasonably common in IP networks, there are many ways with which local rerouting can be provided for high accessibility but most of them are for single link failure. Here, we are suggesting a Localized On-demand Link State (LOLS) routing to safeguard the forwarding continuity even with multiple failures restraining the propagation of failure information to just a few hops. LOLS cannot promise loop-free forwarding during route convergence and this is the motive why we are working on integrating Fast Convergence Fast Reroute (FCFR) practice with LOLS to ensure loop free rerouting and convergence even with many failures. Fast Convergence with Fast Reroute (FCFR), which uses a fast reroute design such as Not-Via and desires just one additional bit in the packet header with much less per-packet overhead. Integrating LOLS with FCFR is going to hold the excellence of LOLS of loop free forwarding and conquer downside of LOLS by ensuring loop free convergence.

Shrinking the hypervisor one subsystem at a time

Efficient and secure networking between virtual machines is crucial in a time where a large share of the services on the Internet and in private datacenters run in virtual machines. To achieve this efficiency, virtualization solutions, such as Qemu/KVM, move toward a monolithic system architecture in which all performance critical functionality is implemented directly in the hypervisor in privileged mode. This is an attack surface in the hypervisor that can be used from compromised VMs to take over the virtual machine host and all VMs running on it. We show that it is possible to implement an efficient network switch nfor virtual machines as an unprivileged userspace component running in the host system including the driver for the upstream network adapter. Our network switch relies on functionality already present in the KVM hypervisor and requires no changes to Linux, the host operating system, and the guest. Our userspace implementation compares favorably to the existing in-kernel implementation with respect to throughput and latency. We reduced per-packet overhead by using a run-to-completion model an are able to outperform the unmodified system for VM-to-VM traffic by a large margin when packet rates are high.

High-Performance Capabilities for 1-Hop Containment of Network Attacks

Capabilities-based networks present a fundamental shift in the security design of network architectures. Instead of permitting the transmission of packets from any source to any destination, routers deny forwarding by default. For a successful transmission, packets need to positively identify themselves and their permissions to the router. A major challenge for a high-performance implementation of such a network is an efficient design of the credentials that are carried in the packet and the verification procedure on the router. We present a capabilities system that uses packet credentials based on Bloom filters. The credentials are of fixed length (independent of the number of routers that are traversed by the packet) and can be verified by routers with a few simple operations. This high-performance design of capabilities makes it feasible that traffic is verified on every router in the network, and most attack traffic can be contained within a single hop. We present an analysis of our design and a practical protocol implementation that can effectively limit unauthorized traffic with only a small per-packet overhead.

Smart codec: an adaptive packet data link

An adaptive convolutional code-rate system for packet data is presented which adapts the level of coding to overcome the effect of varying levels of noise. The system utilises the frame structure of packet data to synchronise adaption and uses a parallel decoding technique to ensure a hitless changeover between a number of different coding rates. The issues concerning the design are discussed in detail, including the resynchronisation of the Viterbi decoders, and the estimation of the level of noise. The system has been implemented using off-the-shelf components and tested over a range of error conditions and introduces only a low per-packet overhead. One important application is to counter the effect of a rain fade on a satellite communication link. A considerable improvement in the link throughput has been observed for a satellite link which suffers from typical rain fades.