Automating cross-layer diagnosis of enterprise wireless networks

Abstract

Modern enterprise networks are of sufficient complexity that even simple faults can be difficult to diagnose -let alone transient outages or service degradations. Nowhere is this problem more apparent than in the 802.11-based wireless access networks now ubiquitous in the enterprise. In addition to the myriad complexities of the wired network, wireless networks face the additional challenges of shared spectrum, user mobility and authentication management. Not surprisingly, few organizations have the expertise, data or tools to decompose the underlying problems and interactions responsible for transient outages or performance degradations. In this paper, we present a set of analysis techniques and models to precisely determine all sources of data transfer delay due to media access and mobility in 802.11 networks -from the physical layer to the transport layer -as well as the interactions among them. While some sources of delay can be directly measured, many of the delay components, such as AP queuing, backoffs, contention, etc., must be inferred. To infer these delays from measurements, we develop a detailed model of MAC protocol behavior, both as it is described in the 802.11 specification as well as how it is implemented in vendor hardware. Combined with comprehensive traces of wireless activity taken from an enterprise network, we produce a complete delay breakdown for packet transmissions and pinpoint problems that constrain connectivity or limit performance.

Abrupt ending of 802.11 AP connections

“…Anomaly detection is such an important aspect in many application domains and has been researched extensively within diverse research areas of IP networks [6][7], wireless 802.11 networks [8] [9][10] and intrusion detection for cyber security [11] [12]. The importance of anomaly detection being in 802.11 networks or in any other application domain is that, usually anomalies in usage data translates to significant, and often critical, actionable decisions.…”

RxIP: Monitoring the health of home wireless networks

“…Centralized EWLAN management has been considered in the context of fault diagnosis [7], protocol extensibility [18], security enhancements, such as detecting rogue APs [4], AP channel assignment and power control [1], client association [19], and client localization [6]. Centaur [24] and Shuffle [16] consider conflictbased per-packet link scheduling, allowing hidden terminal mitigation via scheduling.…”

“…The overhead of coordination is offloaded to the out-of-band wired infrastructure, freeing the wireless spectrum for productive data communication. Deployment experiences show reduced hidden/exposed terminals [7], [24], greater spatial reuse [16], [24], smarter association [19], and a host of other enhancements to the end-user experience [1], [4], [6]. These techniques have proven practical, with commercial systems available from Aruba, Cisco, and Meru [17].…”

WiserAnalyzer: A Passive Monitoring Framework for WLANs

“…among wireless hosts and utilize shared information to analyze and resolve network problems. Cheng et al [17] developed a set of modelling techniques to model delays from physical layer to transport layer and how these delays affect each other. These models can be further used for automatic characterization of causes of the wireless network outages and service degradation.…”