Authors:
Ákos Milánkovich
;
Gergely Ill
;
Károly Lendvai
;
Sándor Imre
and
Sándor Szabó
Affiliation:
Budapest University of Technology and Economics, Hungary
Keyword(s):
Wireless Sensor Networks, Aggregation, Energy Efficiency, FEC, Petri Nets.
Related
Ontology
Subjects/Areas/Topics:
Aggregation, Classification and Tracking
;
Data Manipulation
;
Energy Efficiency
;
Energy Efficiency and Green Manufacturing
;
Industrial Engineering
;
Informatics in Control, Automation and Robotics
;
Network Performance
;
Obstacles
;
Sensor Networks
Abstract:
Energy efficiency is one of the key issues of wireless sensor networks. Aggregation of packets may increase significantly the lifetime of batteries in exchange for some variations in delay. In this paper we have investigated how to determine the optimal amount of packets gathered for aggregation that minimizes the energy consumption of the whole multi-hop network assuming predefined boundary conditions for the delay. To achieve this goal, appropriate models were created to calculate the energy consumption and delay, where we exploited the modelling capabilities of generalized stochastic Petri nets. Using these models, the impact of aggregation was analysed for various test cases. We examined how a network behaves in case of ideal, low and high BERs and investigated how different FEC coding schemes influence the energy consumption. Based on these results, we evaluated the properties of aggregation. We will show, that in case of a good quality radio channel (with low BER) it is not re
commended to use FEC codes to optimize for energy consumption. In case of high aggregation numbers and high BER without the use of FEC the consumed energy converges to infinity. The simulation results show that using the delay as a constraint can narrow down the search for the minimal energy consumption of aggregation number vectors.
(More)