Direct Code Execution Research Articles

On the Fidelity of NS-3 Simulations of Wireless Multipath TCP Connections.

The multipath transmission control protocol (MPTCP) is considered a promising wireless multihoming solution, and the 3rd generation partnership project (3GPP) includes it as a standard feature in the fifth-generation (5G) networks. Currently, ns-3 (Network Simulator-3) is widely used to evaluate the performance of wireless networks and protocols, including the emerging MPTCP protocol. This paper investigates the fidelity of the Linux kernel implementation of MPTCP in the ns-3 direct code execution module. The fidelity of MPTCP simulation is tested by comparing its performance with a real Linux stack implementation of MPTCP using a hardware testbed for two different setups. One setup emulates the existence of a bottleneck link between the sending and receiving networks, whereas the other setup does not have such a bottleneck. The fidelity of ns-3’s simulation is tested for four congestion control algorithms, namely Cubic, linked-increases algorithm (LIA), opportunistic LIA (OLIA) and wVegas for relatively short and long data flows. It is found that the uplink MPTCP throughput performance exhibited by the ns-3 simulator matches the hardware testbed results only if the flows are long-lived and share no common bottleneck link. Likewise, the MPTCP throughput achieved during a downlink scenario using the ns-3 simulator and the hardware testbed are close to each other across all algorithms except wVegas regardless of the flow size if there is no bottleneck link. Moreover, it is observed that the impact of LTE handover on MPTCP throughput is less significant in the simulator than the real hardware testbed, and it is setup-dependent.

Modeling MPTCP Performance

Multi-path TCP (MPTCP) has been standardized by IETF and being used in practice. However, the MPTCP performance analysis has been paid less attention than algorithm improvement, such as coupled congestion control between subflows. The existing analytical models lack consideration of both slow start and congestion avoidance phases. In this letter, we perform a full analysis to derive the MPTCP performance by considering both the phases in the network with shared bottleneck so that we can calculate data transfer latency as well as throughput for a given size of data. Via direct code execution on top of the NS-3 simulation, we show that the proposed analysis captures data transfer latency under packet losses due to buffer overflow.

SIMULACIÓN EN NS3 DEL PROBLEMA DENOMINADO CUELLO DE BOTELLA COMPARTIDO QUE SE PRESENTA EN EL PROTOCOLO MP-TCP

MP-TCP (MultiPath-TCP) is a protocol that allows the sending of data through multiple paths in devices that have several network interfaces. An MP-TCP connection is divided into several TCP connections called subflows, in this way the effective connection rate and the fault resistance can be increased. However, in a scenario where MP-TCP is used with a client and a server, each with a single network interface, you can have the created subflows follow the same path even if several are available; this generates the problem called shared bottleneck. The article describes the development of the C ++ code to simulate the shared bottleneck problem in the NS3 network simulator. The simulation is developed with the DCE (Direct Code Execution) infrastructure for NS3 in such a way that an existing implementation of MP-TCP installed in a Linux kernel can be used. With the developed implementation, tests are performed and the results obtained are presented and discussed, which makes it possible to complement the study and analysis of this problem that MP-TCP suffers from. Among the results, the estimation of the effective transfer rate obtained by varying the number of subflows and roads in scenarios in which the shared bottleneck problem occurs is highlighted.

Dynamic Unicast: Information-centric multi-hop routing for mobile ad-hoc networks

Information-centric networking (ICN) offers new perspectives on mobile ad-hoc communication because routing is based on names but not on endpoint identifiers. Since every content object has a unique name and is signed, authentic content can be stored and cached by any node. If connectivity to a content source breaks, it is not necessarily required to build a new path to the same source but content can also be retrieved from a closer node that provides the same content copy. For example, in case of collisions, retransmissions do not need to be performed over the entire path but due to caching only over the link where the collision occurred. Furthermore, multiple requests can be aggregated to improve scalability of wireless multi-hop communication. In this work, we base our investigations on Content-Centric Networking (CCN), which is a popular ICN architecture. While related works in wireless CCN communication are based on broadcast communication exclusively, we show that this is not needed for efficient mobile ad-hoc communication. With Dynamic Unicast requesters can build unicast paths to content sources after they have been identified via broadcast. We have implemented Dynamic Unicast in CCNx, which provides a reference implementation of the CCN concepts, and performed extensive evaluations in diverse mobile scenarios using NS3-DCE, the direct code execution framework for the NS3 network simulator. Our evaluations show that Dynamic Unicast can result in more efficient communication than broadcast communication, but still supports all CCN advantages such as caching, scalability and implicit content discovery.

DCE: Test the real code of your protocols and applications over simulated networks

We present the Direct Code Execution (DCE) environment for ns-3, notable for being the first free open source framework for integrating Linux kernel and application code within a leading discrete-event network simulator. This new approach has many potential advantages over virtual-machine-based frameworks in terms of realism, reproducibility, avoidance of realtime execution constraints, configuration management, and the ability to debug a network-wide experiment from a single address space using common debugging tools. We provide an overview of DCE and illustrate some key features of this framework with two use cases, one involving thttpd, an HTTP server implementation, and another one involving the udp-perf traffic generator.