Georg Lukas

Precise Admission Control for Bandwidth Reservation in Wireless Mesh Networks

Quality of service in wireless mesh networks is an often requested feature for various kinds of applications. A common approach is the establishment of routes based on hop-by-hop reservation of bandwidth. In this paper we address the problems of admission control in wireless mesh networks. We show that the existing solutions suffer from temporal inconsistencies during the distributed admission process, which lead to high admission failure rates of approximately 2% to 10% in typical topologies, even with the best approach. A solution for this problem is presented, based on the two phase commit protocol. It prevents inconsistencies and the corresponding admission failures.

Modeling Fast Link Failure Detection for Dependable Wireless Mesh Networks

Wireless Mesh Networks (WMN) provide a flexible and inexpensive way to expand network applications to mobile entities. However, high communication reliability demands can not be met by regular WMNs due to mobility and inherent properties of wireless communication. Failing links are common events which impose a major challenge for providing a dependable communication service. By the use of suitable cross-layer models, we provide a timely detection of failed links without compromising the network stability. It is shown that the packet loss caused by a link failure can be reduced from a time-dependent value to at most four consecutive packets. For a typical control application this means an improvement by a factor of 100. In addition to an empirical evaluation we demonstrate the applicability of the mechanism by utilizing it in a real-world WMN to allow the tele-operation of a mobile robot.

Modeling Medium Utilization for Admission Control in Industrial Wireless Mesh Networks

Wireless Mesh Networks (WMNs) are a promising technology for industrial environment communication because of their high flexibility and low cost. To fulfill the thereby arising reliability demands, admission control can be used to prevent congestion and to provide end-to-end guarantees to applications. However, this requires a very precise modeling of the medium utilization for existing and requested data flows, which is challenging due to the dynamics of such networks. In this paper we propose a medium utilization model for small-scale multi-rate WMNs considering the end-to-end throughput. The model is integrated into a feedback control loop to allow an admission control mechanism to ensure reliable end-to-end communication. Our approach increases the available performance without sacrificing reliability. A real test-bed evaluation shows that our admission control manager allows to fully utilize the network capacity while keeping packet losses under 0.5% and three-hop latency below 5.5ms. This is a significant improvement towards reliable multi-rate WMNs, leveraging their applicability in industrial applications with high throughput demands.

Wireless Mesh Network infrastructure for industrial applications — A case study of tele-operated mobile robots

The deployment of wireless communication networks in industrial facilities allows for increased flexibility and seamless integration of mobile participants. Time-critical control applications are generally realized with dedicated wireless sensor networks while common WLAN infrastructure networks allow high-throughput best-effort communication. To provide high-capacity communication as well as real-time service quality, we developed and implemented a Wireless Mesh Network (WMN) infrastructure for industrial applications. An example application that requires these properties is the video-based tele-operation of a mobile robot, which we present as case study in this work. It shows the applicability of our infrastructure as well as the necessity of its comprised components.

Data-Mining-Based Link Failure Detection for Wireless Mesh Networks

Mobile robot applications operating in wireless environments require fast detection of link failures in order to enable fast repair. In previous work, we have shown that cross-layer failure detection can reduce failure detection latency significantly. In particular, we monitor the behavior of the WLAN MAC layer to predict failures on the link layer. In this paper, we investigate data mining techniques to determine which parameters, i.e., the events, or combination and timing of events, occurring on the MAC layer most probably lead to link failures. Our results show, that the parameters revealed with the data mining approach produce similar or even more accurate failure predictions than achieved so far.

Measurement-Based Detection of Interfering Neighbors for QoS in Wireless Mesh Networks

Communication in wireless mesh networks based on the IEEE 802.11 WLAN standard is mainly governed by the carrier sensing based medium access. Knowledge about, which nodes influence each other, can improve the performance and is essential for QoS provision in terms of bandwidth guarantees. However, until now only approximations for the determination of station in carrier sense range are used. We present an exact solution by measuring the carrier sense in static wireless mesh networks. Simulation studies and measurements are done that verify the correctness of the protocol and reveal some significant properties of the carrier sense. It is shown that the carrier sense relation is neither strict nor symmetric in the general case, in opposite to the assumptions normally found in literature. We further conclude that for evaluation better simulation models are required that match these properties.

WMNSec: security for wireless mesh networks

Wireless Mesh Networks (WMNs) are gaining popularity as a flexible and inexpensive replacement for Ethernet-based infrastructure. However, WMN security has not been covered adequately by existing standards and implementations. We propose WMNSec, an adaptation of the IEEE 802.11i security standard, specifically targeted at Wireless Mesh Networks and accounting for limited CPU power, node mobility and interruption-free connectivity. WMNSec has been implemented on top of the MadWifi Linux driver and the hostapd suite. Experimental results from a real WMN show that even in a small eight-node network, WMNSec reduces the authentication time by up to a factor of 3 compared to 802.11i, while allowing mobile stations to move without performing additional authentications. The reduced overhead and the mobility feature confirm the practical usability of WMNSec, finally allowing to deploy WMNs in a secure way.

Preventing admission failures of bandwidth reservation in wireless mesh networks

Quality of Service for wireless mesh networks is an often requested feature for various kinds of applications. A common approach is the hop-by-hop reservation of bandwidth for individual routes. In this paper we address the problems of the reservation on a single hop. In previous works we used simulation studies to show that various existing approaches suffer from inconsistencies that lead to admission failures. In this paper, we discuss the reasons for these failures and present a protocol for preventing them. This allows to significantly increase the reliability of established communication links in WMNs.

Interval based off-line clock synchronization for wireless mesh networks

Wireless mesh networks suffer from various problems like congestion or packet collisions. To identify and overcome these problems an exact global view of the communication is required. However, it is not possible to observe the whole network from a single location. Instead, a distributed monitoring is necessary, which has to include clock synchronization. We present a new interval-based algorithm for the off-line synchronization of passively monitored network events. It calculates the worst-case time interval for every event on a global clock, while considering inaccuracies caused by processing jitter and non-uniform clock drifts. The experimental evaluation on a live mesh network shows an accuracy of better than 130μs over a four-hop distance, which is below the minimum transmission time of data packets. Thereby, our algorithm creates a highly precise global view of the network, which allows a detailed diagnosis of wireless mesh networks.

MLCCA — Multi-Level Composability Check Architecture for dependable communication over heterogeneous networks

During the design of complex networked systems, it is cruical to ensure the composability of the deployed applications and network protocols. Special care has to be taken to provide non-functional requirements like bandwidth and latency. Existing solutions only tackle this problem during the design phase; later refactoring or added components are not covered, potentially causing QoS violations. We propose MLCCA, a multi-level architecture which complements the design-time composability checks with additional automatic checks performed at compiletime and at run-time. The required infrastructure is embedded into our communication middleware FAMOUSO, making it transparent to application developers. The architecture has been evaluated in a tele-operated mobile robot case study. If the QoS attributes could not be fulfilled due to refactoring or changed conditions, no communication was allowed by the middleware, ensuring that the application could enter a fail-safe state. No data was sent over insufficient channels. Thus, our combination of FAMOUSO and MLCCA enables the sustainable deployment of complex networked systems.