Testbed

Traffic Monitoring and Classification

KAUMesh has available several traffic generators (such as MGEN, IPERF, etc) which help you in your experiments. You can also port your own application and the monitoring interfaces allow to capture traffic or monitor it. We have customized Nagios platform to allow live moinitoring. Also, we have implemented one way delay monitoring facilities. This allows you to implement tools to monitor traffic in multiradio multichannel meshes.

Bandwidth Estimation

We ported several bandwidth monitoring tools for KauMesh such as Abing and Pathload. You can also implement/port your own tool and evaluate it in a multiradio multichannel setting.

Channel model validation

KAUMesh provides a good opportunity to conduct measurements which can be used to calibrate link and channel models against realistic interference effects under different combinations of frequencies, channel bandwidth, distances between nodes...

Adaptive coding/modulation

An interesting evaluation possibility that mesh testbeds offer is the possibility to select different modulation schemas for different links and channels, combined with different carrier sense thresholds and sending power. Therefore, there is the possibility to evaluate the impact of changing modulation schemas in a multichannel environment.

Network coding

KAUMesh is a multi-hop multichannel multi-radio indoor testbed. As such it is perfectly suited to study the effect of network coding in this kind of environment. It allows to develop and to evaluate novel schemas to optimize e.g. multicast delivery in such networks using network coding. Also, it would be good candidate to develop and evaluate channel assignment algorithms that take into account the opportunity of coded overhearing, etc.

Relaying and cooperation in networks

Due to its multi-hop nature, it requires cooperation among nodes. Therefore, it is very well suited to evaluate the performance of novel estimate-and-forward protocols that are based on joint source-channel coding or even joint source network coding focusing on higher layer and cross-layer aspects. In Virtual Antenna Arrays, it would be very interesting to develop and evaluate novel schemas that make use of superposition coding either for downlink or uplink. In addition, KAUMesh is a very interesting candidate to study the interaction between network layer protocol and cross-layer optimization techniques. Currently, KAUMesh supports and implements both AODV and OLSR routing and includes plug-ins for channel assignment into the routing protocol. In a similar way, it is straightforward with KAUMesh to use additional routing information for the purposes of organizing cooperative relaying at the link layer.

Joint source and channel coding/decoding

We have already tested with KAUMesh effective VoIP and Video streaming over multi-channel multi-radio meshed network from within the mesh to destination located in the Internet. Therefore, the incorporation of UEP for streaming application seems feasible to evaluate using KAUMesh. In what congestion concerns, there is also the possibility to incorporate congestion aware mechanisms in the joint source/channel coding and take this into account in the route selection to e.g. maximize user utility for all flows within the network. As a result, KAUMesh is a good opportunity to evaluate novel video delivery mechanisms in a multi-hop environment.

Scheduling and adaptive radio resource assignment

Due to its self-organizing nature and the possibilities to serve multiple channels and frequency bands in parallel, it would be interesting to develop and evaluate novel scheduling mechanisms for multi-channel mesh networks using KAUMesh as experimental platform.

Joint RRM and flexible use of radio spectrum

In KAUMesh, radios can tune to any of the frequency bands available in 802.11 a/b/g band dynamically with minimum switching overhead (around 1000 µs). In addition, radios can be tuned to 5, 10, 20 and 40 MHz. Modulation schemas can be changed as well as sending power and carrier sense threshold. This opens up the opportunity to test and evaluate novel RRM mechanisms and study the effect of e.g. different channel bandwidth on interference and performance. When it comes to channel assignment, currently we use a simple strategy to balance the number of neighbors on a channel. However, more advanced mechanisms can be designed, implemented and evaluated through KAUMesh based on e.g. Game Theoretical Approaches. Such approaches can be combined with cognition capabilities so that the MAC layer can learn better adaptation strategies over time (e.g. which frequency to use, what adaptations of parameters of PHY layer to select, etc.)

Opportunistic networks

As mesh testbeds can be deployed as two different not-connected islands, they can be used to evaluate delay tolerant networking protocols utilizing peoples mobility or scheduled links (such as a wireless node placed on an elevator to opportunistically connect those two islands) to disseminate data. In addition, mesh testbeds can serve as a mechanisms to evaluate opportunistic routing protocols.