International Conference Papers

Permanent URI for this collectionhttp://dl.cerist.dz/handle/CERIST/4

Browse

Author: search.filters.author.Djenouri, DjamelSubject: search.filters.subject.Energy-Efficiency

Search Results

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    BA: Game Theoretical Approach for Energy-Delay Balancing in Distributed Duty-Cycled MAC Protocols of Wireless Networks
    (ACM, 2014-07-14) Doudou, Messaoud; M. Barcelo-Ordinas, Jose; Djenouri, Djamel; Garcia-Vidal, Jorge; Badache, Nadjib
    Optimizing energy consumption and end-to-end (e2e) packet delay in energy constrained distributed wireless networks is a conflicting multi-objective optimization problem. This paper investigates this trade-off from a game-theoretic perspective, where the two optimization objectives are considered as virtual game players that attempt to optimize their utility values. The cost model of each player is mapped through a generalized optimization framework onto protocol specific MAC parameters. A cooperative game is then defined, in which the Nash Bargaining solution assures the balance between energy consumption and e2e packet delay. For illustration, this formulation is applied to three state-of-the-art wireless sensor network MAC protocols; X-MAC, DMAC, and LMAC as representatives of preamble sampling, slotted contention-based, and frame-based MAC categories, respectively. The paper shows the effectiveness of such framework in optimizing protocol parameters for achieving a fair energy-delay performance trade-off, under the application requirements in terms of initial energy budget and maximum e2e packet delay. The proposed framework is scalable with the increase in the number of nodes, as the players represent the optimization metrics instead of nodes.
  • Thumbnail Image
    Item
    Cost Effective Node Deployment Strategy for Energy-Balanced and Delay-Efficient Data Collection in Wireless Sensor Networks
    (2014) Doudou, Messaoud; Djenouri, Djamel; M. Barcelo-Ordinas, Jose; Badache, Nadjib
    The real-world node deployment aspect is investigated, while considering cost minimization for resolving the energy hole around the sink, which represents a serious problem in typical sensor networks with uniform distribution. A novel strategy is proposed that is based on the use of two sinks and a few extra relay nodes close to the sinks' areas. The traffic is then alternatively sent to the sinks in every other cycle. As a second contribution, an efficient data collection mechanism has been developed to determine the optimal data rate that meets delay requirements of sensors and improves the network lifetime. The comparison of the proposed node deployment strategy with uniform, non-uniform geometric and linear increase node distributions demonstrates that the cost of the proposed solution is very close to that of the uniform distribution and much lower than others, while achieving a load balancing at the same order of the state-of-the-art solutions.