Medium Access Control Protocols in Mobile Ad-hoc Networks Using Beaconing

Abstract

In Mobile Ad-hoc Networks using Beaconing (beaconing MANETs), network nodes periodically broadcast beacons about their status. Transportation systems rely on this type of networks, among others for collision avoidance systems. In comparison to traditional MANETs, Medium Access Control (MAC) protocols face new challenges, as broadcast is used and the network topology is unknown and subject to highly dynamic changes. These challenges are addressed by a meaningful metric for performance analysis (the update delay), a methodology to fairly compare, to predict and to optimise the performance of MAC protocols for a given application and a novel MAC protocol suitable for beaconing MANETs, COMB. The work is validated with a novel, infrastructureless collision avoidance system, RCAS, for which this work suggests a two-layered bio-inspired approach. The update delay measures the time elapsed in a receiver between two consecutive beacons of a sender. In contrast to the collision rate, the complementary cumulative distribution function of the update delay can detect e.g. for CSMA the negative impact of the periodic beaconing and back-off strategy. Furthermore, it has shown that SOTDMA only works ideally under the conditions of a closed network. In an open network, its performance degrades almost to the one of Slotted Aloha or even worse if the size of the necessary MAC header of SOTDMA is not negligible. For realistic performance prediction, the application, system and environmental parameters are broken down into generic MAC performance factors: observed offered traffic, the MAC situational information, hidden nodes, beaconing frequency, renewal cycle, capture effect, interference and specific protocol parameters. For protocol optimisation, also the application requirements have to be deduced correctly, the worst-case scenario, the maximum required update delay and safety level as well as the communication parameters. Knowing the effect of these factors, the COMB protocol has been developed. This two-layered protocol divides the communication network in cells using different channels which represent virtually closed networks as they are smaller than the intended communication range. Channel separation (upper channel) and medium access in the single channels (lower channel) can be realised by different strategies. The evaluations show that SOTDMA is the best option for the lower level of COMB. This combination outperforms the state of the art with Slotted Aloha, SOTDMA and CSMA by the 100-fold. With the concepts and results in this thesis, application providers in beaconing MANETs can identify the MAC protocol ideal for their application without the need for extensive simulations or unrealistic simplifications.