TY - ECHAP
KW - History-Based Prediction
KW - Mobile Ad Hoc Networks
KW - Mobile Collaboration
KW - Network Topology Prediction
KW - Routing Protocols
AU - Pere Millan
AU - Carlos Molina
AU - Roc Meseguer
AU - Sergio Ochoa
AU - Rodrigo Santos
AU - Giancarlo Fortino
AU - Giuseppe Di Fatta
AU - Wenfeng Li
AU - Sergio Ochoa
AU - Alfredo Cuzzocrea
AU - Mukaddim Pathan
AB - Several social computing participation strategies, such as crowdsensing and crowdsourcing, use mobile ad hoc or opportunistic networks to support the users activities. The unreliability and dynamism of these communication links make routing protocols a key component to achieve efficient and reliable data communication in physical environments. Often these routing capabilities come at expenses of flooding the network with a huge amount of topology control information (TCI), which can overload the communication links and dramatically increase the energy consumption of the participating devices. In previous works the authors have shown that predicting the network topology in these work scenarios helps reduce the number of control packets delivered through the network. This saves energy and increases the available bandwidth. This paper presents a study that extends the authors’ previous works, by identifying the impact of predicting the TCI generated by routing protocols in these networks. The prediction process is done following a history-based approach that uses information of the nodes past behavior. The paper also determines the predictability limits of this strategy, assuming that a TCI message can be correctly predicted if it appeared at least once in the past. The results show that the upper-bound limit of the history-based prediction approach is high, and that realistic prediction mechanisms can achieve significant ratios of accuracy. Mobile collaborative applications and routing protocols using mobile ad hoc or opportunistic networks can take advantage of this prediction approach to reduce network traffic, and consequently, the energy consumption of their devices.
DO - 10.1007/978-3-319-11692-1_21
N2 - Several social computing participation strategies, such as crowdsensing and crowdsourcing, use mobile ad hoc or opportunistic networks to support the users activities. The unreliability and dynamism of these communication links make routing protocols a key component to achieve efficient and reliable data communication in physical environments. Often these routing capabilities come at expenses of flooding the network with a huge amount of topology control information (TCI), which can overload the communication links and dramatically increase the energy consumption of the participating devices. In previous works the authors have shown that predicting the network topology in these work scenarios helps reduce the number of control packets delivered through the network. This saves energy and increases the available bandwidth. This paper presents a study that extends the authors’ previous works, by identifying the impact of predicting the TCI generated by routing protocols in these networks. The prediction process is done following a history-based approach that uses information of the nodes past behavior. The paper also determines the predictability limits of this strategy, assuming that a TCI message can be correctly predicted if it appeared at least once in the past. The results show that the upper-bound limit of the history-based prediction approach is high, and that realistic prediction mechanisms can achieve significant ratios of accuracy. Mobile collaborative applications and routing protocols using mobile ad hoc or opportunistic networks can take advantage of this prediction approach to reduce network traffic, and consequently, the energy consumption of their devices.
PB - Springer International Publishing
PY - 2014
SN - 978-3-319-11691-4
SP - 237
EP - 249
TI - Using a History-Based Approach to Predict Topology Control Information in Mobile Ad Hoc Networks
UR - http://dx.doi.org/10.1007/978-3-319-11692-1_21
VL - 8729
ER -